How To Calculate Fractional Crystallization

Estimate how trace element concentration changes in melt and crystals during fractional crystallization.

How to Calculate Fractional Crystallization: A Practical Expert Guide

Fractional crystallization is one of the most important processes in igneous petrology. It explains how a single parent magma can evolve into multiple daughter compositions as minerals crystallize and are separated from the melt. If you are trying to learn how to calculate fractional crystallization, the key is understanding the relationship between melt fraction (F), partition coefficient (D), and the concentration of a chemical component in the evolving liquid and solids.

Why these calculations matter

In natural systems, magma does not stay compositionally static. As temperature drops, minerals like olivine, pyroxene, plagioclase, and Fe-Ti oxides start to crystallize in sequence. Those crystals remove specific elements from the liquid, depending on each element’s compatibility with those minerals. This drives the residual melt toward a different composition. Geoscientists use quantitative fractional crystallization models to:

• Reconstruct magmatic histories in volcanic arcs, rifts, and mid-ocean ridge settings.
• Estimate how strongly trace elements become enriched or depleted during cooling.
• Compare measured rock compositions to theoretical differentiation paths.
• Test hypotheses about crustal contamination versus closed-system evolution.

When you compute these trends correctly, you can bridge field observations, petrography, and geochemical data into one coherent interpretation.

The core equation: Rayleigh fractional crystallization

The most widely used first-order model is Rayleigh fractional crystallization for a closed system with perfect crystal removal. The liquid concentration of an element is:

C_l = C₀ × F^{(D – 1)}

Where:

• C_l = concentration in the remaining melt.
• C₀ = initial concentration in the parent melt.
• F = mass fraction of melt remaining (0 to 1).
• D = bulk partition coefficient for the crystallizing assemblage.

The concentration in the instantaneous crystal that forms at a given step is:

C_s,inst = D × C_l

The average concentration in all solids removed up to that point is often estimated by:

C_s,avg = C₀ × (1 – F^D) / (1 – F)

Step-by-step method to calculate fractional crystallization

1. Choose your element and initial concentration: for example, Sr = 300 ppm in parental basalt.
2. Estimate bulk D: based on your mineral assemblage and P-T conditions. For Sr with abundant plagioclase, D can be closer to or above 1. For highly incompatible elements, D is usually much less than 1.
3. Define F: if 45% melt remains, F = 0.45.
4. Apply the Rayleigh formula: calculate C_l from C₀, D, and F.
5. Compute crystal concentrations: calculate instantaneous and average solid concentrations.
6. Check mass balance: if initial melt mass is known, convert concentrations to total element mass in liquid and solids.

Interpretation rule of thumb: if D is less than 1, the element is incompatible and becomes progressively enriched in the remaining melt as F decreases. If D is greater than 1, the element is compatible and tends to be removed into crystals, depleting the liquid.

Typical partition behavior statistics used in calculations

Bulk D values are not fixed constants. They vary with pressure, temperature, oxygen fugacity, and mineral modes. Still, geochemists often begin modeling with experimentally constrained ranges. The table below compiles commonly used ranges for basaltic to basaltic-andesite differentiation studies.

Element	Typical Bulk D Range	Compatibility Class	Practical Modeling Impact
Rb	0.01 to 0.10	Strongly incompatible	Rapid enrichment in residual melt at low F
Ba	0.05 to 0.30	Incompatible	Strong liquid enrichment unless K-feldspar dominates late stages
La	0.02 to 0.20	Incompatible REE	High sensitivity to small amounts of crystallization
Sr	0.5 to 2.0	Variable, often compatible with plagioclase	Can switch from mild enrichment to strong depletion depending on mineralogy
Ni	2 to 10+	Compatible in olivine/pyroxene systems	Strong depletion in evolved liquids
Cr	5 to 50+	Strongly compatible with spinel/pyroxene	Very fast drop in melt concentration during early crystallization

These ranges are consistent with common experimental and field-based petrology datasets and are widely used for first-pass Rayleigh modeling before calibration to specific local rock suites.

How magma composition evolves with crystallization

Fractional crystallization does not only affect trace elements. It also drives whole-rock differentiation from mafic to more silicic compositions. As a generalized trend, basaltic melts can evolve toward andesite, dacite, and rhyolite as Fe-Mg-rich minerals are removed and silica activity in the liquid rises.

Magma Type	Typical SiO2 (wt%)	Typical Eruption Temperature (°C)	Viscosity Trend
Basalt	45 to 52	1000 to 1200	Lowest viscosity among common magmas
Andesite	52 to 63	800 to 1000	Intermediate viscosity
Dacite	63 to 69	750 to 900	High viscosity
Rhyolite	69 to 77+	650 to 800	Very high viscosity, often explosive behavior

These composition and temperature ranges are consistent with commonly reported igneous classifications used by geological surveys and university petrology curricula.

Worked conceptual example

Suppose a parent magma contains 300 ppm of an incompatible element and has a bulk D of 0.2. If melt fraction drops to F = 0.45, then:

• C_l = 300 × 0.45^-0.8 ≈ enriched relative to the parent melt.
• C_s,inst = 0.2 × C_l, so new crystals have lower concentration than melt.
• As F decreases further, liquid enrichment accelerates nonlinearly.

Now compare with a compatible element where D = 2.5. At the same F, C_l would be significantly depleted. This contrast is the core reason why incompatible trace element ratios are powerful indicators of fractionation progress.

Common mistakes when calculating fractional crystallization

1. Mixing percent and fraction for F: entering 45 instead of 0.45 is a major error.
2. Using unrealistic D values: always anchor D in the expected mineral assemblage.
3. Ignoring changing mineral modes: bulk D may evolve through time.
4. Comparing unlike units: keep ppm, ppb, and wt% consistent.
5. Forgetting open-system effects: recharge, assimilation, and mixing can overprint Rayleigh trends.

Advanced interpretation tips

If you need higher confidence interpretation, combine this calculator with petrographic constraints and multiple element models. For example, fit incompatible element trends (Rb, La, Nb) together with compatible element depletion (Ni, Cr). If one set fits but the other does not, the system may involve crystal accumulation or magma mixing instead of pure fractional crystallization.

You can also run sensitivity tests by varying D and F to generate envelopes rather than a single curve. In real magma chambers, uncertainties in D can be larger than analytical uncertainty in concentrations, so scenario ranges are often more meaningful than one best-fit value.

Authoritative learning resources

For deeper background and verified educational material, consult:

• U.S. Geological Survey (USGS) Volcano Hazards Program
• USGS overview of Bowen’s Reaction Series
• Carleton College geoscience educational resources (.edu)

These sources are useful for linking numerical modeling to mineral crystallization sequences, petrologic equilibrium concepts, and volcanic system behavior.