Atomic Structure, the Periodic Table and Ionic Substitution
CSIR NET Earth Sciences geochemistry: nuclide stability, abundances, ionic radii, radius ratio, Goldschmidt and Ringwood rules, LILE, HFSE and REE behaviour.
- 19 min
On this page · 13 sections
- Prerequisites
- The Hook: Why Nickel Stays in the Mantle and Rubidium Ends Up in Granite
- The Atom for a Geochemist
- Ionic Radius: the Master Variable
- Ionic Substitution in Minerals
- Classifying Trace Elements by Behaviour
- Memory Tricks
- Don’t Confuse
- Common Mistakes
- Practice
- Quick Revision Table
- Key Takeaways
- What’s Next
Prerequisites
- Electron configuration, effective nuclear charge and periodic trends at MSc chemistry level. If they are rusty, read Electronic Configuration and Effective Nuclear Charge and Atomic and Ionic Radii.
- Nuclear stability and decay modes: Nuclear Stability and Radioactive Decay.
- Goldschmidt’s siderophile–chalcophile–lithophile–atmophile classes, introduced in Part B: Meteorites, Earth’s Composition and Differentiation.
The Hook: Why Nickel Stays in the Mantle and Rubidium Ends Up in Granite
The Atom for a Geochemist
Nucleus: what makes an isotope
An atom of element X is written \( ^{A}_{Z}\text{X} \): \( Z \) protons fix the element and its chemistry, \( N \) neutrons fix the isotope, and \( A = Z + N \) is the mass number.
| Term | Same | Different | Example |
|---|---|---|---|
| Isotopes | \( Z \) | \( N \), \( A \) | \( ^{16}\text{O} \), \( ^{17}\text{O} \), \( ^{18}\text{O} \) |
| Isobars | \( A \) | \( Z \) | \( ^{87}\text{Rb} \) and \( ^{87}\text{Sr} \); \( ^{40}\text{K} \), \( ^{40}\text{Ar} \), \( ^{40}\text{Ca} \) |
| Isotones | \( N \) | \( Z \), \( A \) | \( ^{13}\text{C} \) and \( ^{14}\text{N} \) (\( N = 7 \)) |
Isobars matter in geochemistry because most radiogenic pairs are isobars: a \( \beta^- \) decay changes \( Z \) by one and leaves \( A \) unchanged. That is why a mass spectrometer must chemically separate Rb from Sr before measuring \( ^{87}\text{Sr} \).
Nuclear stability. Stable nuclides lie in a narrow band on a plot of \( N \) against \( Z \): \( N \approx Z \) for light nuclides, rising to \( N/Z \approx 1.5 \) for the heaviest. Nuclei with even \( Z \) and even \( N \) are the most numerous stable nuclides; odd–odd stable nuclides are very few (\( ^{2}\text{H} \), \( ^{6}\text{Li} \), \( ^{10}\text{B} \), \( ^{14}\text{N} \) and a handful of others). Nuclides off the band decay towards it (radiogenic isotopes, treated in Radiogenic Isotopes and Geochronology).
Cosmic abundances: the pattern behind Earth’s chemistry
The solar-system abundance curve (from the solar photosphere and CI chondrites) has four features that Part C reuses:
- H and He dominate, together about 98% of the mass; abundance falls roughly exponentially with \( Z \) up to the iron peak.
- Li, Be and B are strongly depleted: they are destroyed in stellar interiors and made mainly by cosmic-ray spallation.
- An iron peak near \( A = 56 \): nuclear binding energy per nucleon is highest around Fe and Ni, so stellar fusion ends there.
- The Oddo–Harkins rule: elements with even \( Z \) are more abundant than their odd-\( Z \) neighbours, giving a saw-tooth curve. This is why REE data are always divided by chondrite before plotting (see below).
Electrons: what fixes chemistry
| Block | Shell being filled | Geochemical consequence |
|---|---|---|
| s (groups 1, 2) | ns | form \( \text{M}^+ \), \( \text{M}^{2+} \) with noble-gas cores; large, weakly charged, mostly ionic bonds (Na, K, Rb, Cs, Mg, Ca, Sr, Ba) |
| p (groups 13–18) | np | Si, Al, P, S, O, C, halogens; Si–O bond roughly half covalent; noble gases inert (atmophile) |
| d (transition metals) | (n−1)d | variable oxidation state (\( \text{Fe}^{2+}/\text{Fe}^{3+} \), \( \text{Mn}^{2+}/\text{Mn}^{4+} \), \( \text{Cr}^{3+}/\text{Cr}^{6+} \)); crystal-field effects decide site preference |
| f (lanthanides, actinides) | 4f, 5f | REE are almost all \( 3+ \) and shrink steadily (lanthanide contraction); exceptions Eu\(^{2+}\) and Ce\(^{4+}\); U and Th tetravalent in magmas, U hexavalent in oxidised water |
Electronegativity and bond type. Pauling’s estimate of ionic character for a bond between atoms with electronegativity difference \( \Delta\chi \) is
$$\text{ionic fraction} \approx 1 - e^{-0.25(\Delta\chi)^2}$$With \( \chi(\text{O}) = 3.44 \): Si–O (\( \Delta\chi = 1.54 \)) is about 45% ionic, Al–O about 57%, Mg–O about 68%, K–O about 82%. So silicates are neither purely ionic nor purely covalent; the ionic model works best for the large, low-charge cations.
Ionic Radius: the Master Variable
What controls the size of an ion
- Down a group: radius increases (more shells): \( \text{Li}^+ < \text{Na}^+ < \text{K}^+ < \text{Rb}^+ < \text{Cs}^+ \).
- Charge: for one element, higher charge means a smaller ion: \( \text{Fe}^{3+} < \text{Fe}^{2+} \); \( \text{Eu}^{3+} < \text{Eu}^{2+} \).
- Coordination number (CN): the same ion is larger in a higher-CN site. Radii must be compared at the same CN.
- Spin state: high-spin \( \text{Fe}^{2+} \) is larger than low-spin; crustal and upper-mantle minerals carry high-spin Fe.
- Lanthanide contraction: across La to Lu the radius of \( \text{REE}^{3+} \) falls steadily, because 4f electrons shield the growing nuclear charge poorly. Two results: heavy REE (HREE) fit garnet better than light REE (LREE), and Hf has almost exactly the radius of Zr.
Effective ionic radii
Values below are the effective ionic radii of Shannon (1976), in ångström (Å), for the coordination shown in Roman numerals. Use them for comparison, not as exact constants.
| Ion | CN | Radius (Å) | Ion | CN | Radius (Å) |
|---|---|---|---|---|---|
| \( \text{O}^{2-} \) | VI | 1.40 | \( \text{Mg}^{2+} \) | VI | 0.72 |
| \( \text{Si}^{4+} \) | IV | 0.26 | \( \text{Fe}^{2+} \) (high spin) | VI | 0.78 |
| \( \text{Al}^{3+} \) | IV / VI | 0.39 / 0.535 | \( \text{Ni}^{2+} \) | VI | 0.69 |
| \( \text{Fe}^{3+} \) (high spin) | VI | 0.645 | \( \text{Ca}^{2+} \) | VIII | 1.12 |
| \( \text{Ti}^{4+} \) | VI | 0.605 | \( \text{Sr}^{2+} \) | VIII | 1.26 |
| \( \text{Zr}^{4+} \) | VIII | 0.84 | \( \text{Na}^{+} \) | VIII | 1.18 |
| \( \text{Hf}^{4+} \) | VIII | 0.83 | \( \text{K}^{+} \) | VIII | 1.51 |
| \( \text{Li}^{+} \) | VI | 0.76 | \( \text{Rb}^{+} \) | VIII | 1.61 |
| \( \text{Cr}^{3+} \) | VI | 0.615 | \( \text{Ba}^{2+} \) | VIII | 1.42 |
Radius ratio and coordination (Pauling’s first rule)
Around each cation, anions form a coordination polyhedron. The cation must touch its anions, so the stable CN depends on \( r_c/r_a \):
| \( r_c/r_a \) | CN | Polyhedron | Typical cations (with \( \text{O}^{2-} \)) |
|---|---|---|---|
| 0.155–0.225 | 3 | triangle | \( \text{C}^{4+} \) (in \( \text{CO}_3^{2-} \)), \( \text{B}^{3+} \) |
| 0.225–0.414 | 4 | tetrahedron | \( \text{Si}^{4+} \), \( \text{Al}^{3+} \) (part) |
| 0.414–0.732 | 6 | octahedron | \( \text{Mg}^{2+} \), \( \text{Fe}^{2+} \), \( \text{Al}^{3+} \) (part), \( \text{Ti}^{4+} \) |
| 0.732–1.0 | 8 | cube | \( \text{Ca}^{2+} \), \( \text{Na}^{+} \) |
| ≥ 1.0 | 12 | cubo-octahedron | \( \text{K}^{+} \), \( \text{Rb}^{+} \), \( \text{Ba}^{2+} \) |
Worked checks with the radii above: \( \text{Mg}^{2+} \): \( 0.72/1.40 = 0.51 \), octahedral, as in olivine and pyroxene. \( \text{K}^{+} \): \( 1.51/1.40 = 1.08 \), 8 to 12 coordination, as in feldspar and mica. \( \text{Al}^{3+} \): \( 0.535/1.40 = 0.38 \), close to the 0.414 boundary, which is why Al occupies both tetrahedral and octahedral sites (Al\(^{\text{IV}}\) in feldspar, Al\(^{\text{VI}}\) in kyanite and the octahedral sheet of micas).
Ionic potential: who dissolves, who stays
Ionic potential (field strength) \( \phi = Z/r \) measures how strongly a cation attracts the oxygen of water. It sorts elements in weathering and in magmas:
| \( Z/r \) (r in Å) | Behaviour in water | Examples |
|---|---|---|
| Low (below about 3) | stay as hydrated cations; soluble; carried to the sea | Na, K, Ca, Mg, Rb, Sr, Ba |
| Intermediate (about 3–12) | hydrolyse and precipitate as hydroxides; insoluble; stay in soils and bauxite | Al, Fe\(^{3+}\), Ti, Zr, Th |
| High (above about 12) | pull O from water to form soluble oxyanions | C (\( \text{CO}_3^{2-} \)), N (\( \text{NO}_3^- \)), S (\( \text{SO}_4^{2-} \)), P (\( \text{PO}_4^{3-} \)), B |
This is Goldschmidt’s classic ionic-potential diagram. The same idea explains why bauxite and laterite are Al- and Fe-rich residues of intense weathering (see Weathering, Erosion, Transport and Deposition). The boundaries are approximate and depend on the radius set used.
Ionic Substitution in Minerals
Ionic substitution (diadochy, isomorphous replacement) is the replacement of one ion by another of similar size and charge in a crystal site. It produces solid solutions (olivine Fo–Fa, plagioclase Ab–An, garnet) and controls where trace elements go.
Goldschmidt’s rules
- Size: extensive substitution needs radii within about 15% of each other. Beyond that, substitution is limited; beyond about 30%, negligible.
- Charge: substitution is easy when charges differ by at most one unit, provided electrical neutrality is kept by a coupled substitution elsewhere.
- Priority by radius: of two ions with the same charge and similar radius, the smaller enters a growing crystal preferentially.
- Priority by charge: of two ions with similar radius and different charge, the one with the higher charge enters preferentially.
The 15% figure is a rule of thumb, not a law. Temperature widens the tolerance (high-temperature minerals accept more misfit, which is why exsolution lamellae form on cooling).
Ringwood’s electronegativity rule (1955). Of two ions with similar radius and charge, the one with the lower electronegativity is preferred, because it forms a more ionic bond with oxygen. This explains why \( \text{Cu}^{+} \) (close to \( \text{Na}^{+} \) in size in older radius sets, but much more electronegative) does not enter Na sites of feldspar, and why Zn, with a radius close to \( \text{Mg}^{2+} \) and \( \text{Fe}^{2+} \), is less abundant in olivine and pyroxene than size alone predicts and tends to form sulphides and spinels.
Camouflage, capture and admission
Goldschmidt’s terms for how a trace ion relates to the major ion it replaces:
| Term | Trace ion compared with major ion | Result | Examples |
|---|---|---|---|
| Camouflage | same charge, similar radius | concentration in mineral ≈ in melt; no fractionation | \( \text{Hf}^{4+} \) for \( \text{Zr}^{4+} \) in zircon; \( \text{Ga}^{3+} \) for \( \text{Al}^{3+} \); \( \text{Ni}^{2+} \) for \( \text{Mg}^{2+} \) (in size) |
| Capture | higher charge (or smaller radius, same charge) | trace ion enriched in early crystals | \( \text{Ba}^{2+} \) or \( \text{Sr}^{2+} \) for \( \text{K}^{+} \) in K-feldspar |
| Admission | lower charge (or larger radius) | trace ion enriched in the residual melt and late minerals | \( \text{Li}^{+} \) for \( \text{Mg}^{2+} \); \( \text{Rb}^{+} \) for \( \text{K}^{+} \) |
Rb illustrates admission by size: \( \text{Rb}^{+} \) is 6.6% larger than \( \text{K}^{+} \) at CN VIII, so K/Rb falls as a granitic magma evolves and late pegmatites have low K/Rb.
Coupled substitution
When charges differ, two substitutions occur together to keep the crystal neutral:
| Mineral | Coupled substitution | Result |
|---|---|---|
| Plagioclase | \( \text{Na}^{+}\text{Si}^{4+} \rightleftharpoons \text{Ca}^{2+}\text{Al}^{3+} \) | albite–anorthite series |
| Pyroxene (Tschermak) | \( \text{Mg}^{2+}\text{Si}^{4+} \rightleftharpoons \text{Al}^{3+}\text{Al}^{3+} \) | Al in pyroxene, pressure- and temperature-sensitive |
| Pyroxene (jadeite) | \( \text{Ca}^{2+}\text{Mg}^{2+} \rightleftharpoons \text{Na}^{+}\text{Al}^{3+} \) | diopside–jadeite (omphacite in eclogite) |
| Amphibole (edenite) | \( \square + \text{Si}^{4+} \rightleftharpoons \text{Na}^{+} + \text{Al}^{3+} \) | vacancy (\( \square \)) in the A site filled |
| Zircon, xenotime | \( \text{Zr}^{4+} + \text{Si}^{4+} \rightleftharpoons \text{REE}^{3+} + \text{P}^{5+} \) | HREE and Y in zircon |
Crystal-field effects
For transition-metal ions with partly filled d orbitals, the octahedral site in a silicate splits the d levels and lowers the energy of some configurations. The octahedral site preference energy is large for \( \text{Cr}^{3+} \) (d\(^3\)) and \( \text{Ni}^{2+} \) (d\(^8\)) and zero for \( \text{Mn}^{2+} \) and \( \text{Fe}^{3+} \) (high-spin d\(^5\)) and for \( \text{Zn}^{2+} \) (d\(^{10}\)). So Ni and Cr are strongly retained in the octahedral sites of olivine, pyroxene and spinel, far beyond what radius alone predicts; Mn and Zn are not. Details are in Crystal Field Theory, Spectroscopy and Defects.
Classifying Trace Elements by Behaviour
Compatible and incompatible
The partition coefficient \( D = C_{\text{mineral}}/C_{\text{melt}} \) (developed in Distribution Coefficients, Nucleation and Diffusion) sorts elements:
- Compatible (\( D > 1 \)): stay in the mantle residue or early cumulates. Ni and Co (olivine), Cr (spinel, cpx), Sc and V (pyroxene), Sr and Eu (plagioclase), HREE and Y (garnet).
- Incompatible (\( D < 1 \)): concentrate in melts and therefore in the continental crust. Two groups by field strength:
| Group | Ionic potential | Members | Behaviour |
|---|---|---|---|
| LILE (large-ion lithophile, low field strength) | low (below about 2) | K, Rb, Cs, Sr, Ba, Pb, Eu\(^{2+}\) | soluble in aqueous fluids; mobile during alteration, metamorphism and slab dehydration |
| HFSE (high field strength) | high (above about 2) | Zr, Hf, Ti, Nb, Ta, Th, U\(^{4+}\), and the REE in many schemes | insoluble in aqueous fluids; immobile in alteration, so used for tectonic discrimination of altered rocks |
The boundary near \( Z/r = 2 \) is the usual convention. Arc magmas show high LILE/HFSE (for example high Ba/Nb) and a negative Nb–Ta anomaly, because slab fluids carry LILE but not HFSE.
The rare earth elements (REE)
- La to Lu (\( Z = 57 \)–71); Y behaves like Ho (almost identical radius and charge). LREE La–Nd (or La–Sm), MREE Sm–Ho, HREE Er–Lu. Pm has no stable isotope.
- All \( 3+ \) except Eu\(^{2+}\) (reducing magmas; half-filled 4f\(^7\)) and Ce\(^{4+}\) (oxidising surface waters; empty 4f\(^0\)).
- Chondrite normalisation: divide each REE by its value in CI chondrite to remove the Oddo–Harkins saw-tooth and show fractionation as a smooth line. MORB is LREE-depleted; continental crust and OIB are LREE-enriched.
- Eu anomaly: \( \text{Eu}^{2+} \) (radius close to \( \text{Sr}^{2+} \)) enters plagioclase. Plagioclase cumulates show a positive anomaly; melts that crystallised plagioclase show a negative one. With subscript N for chondrite-normalised values,
Values below 1 are negative anomalies.
- Ce anomaly: in oxic seawater, \( \text{Ce}^{3+} \) oxidises to insoluble \( \text{Ce}^{4+} \) and is scavenged onto Fe–Mn oxides, so seawater has a negative Ce anomaly and Fe–Mn nodules a positive one. Ce anomalies in ancient marine sediments are used as palaeo-redox indicators (see Redox, Eh–pH, Waters and Aerosols).
- Garnet signature: garnet strongly retains HREE, so melts formed with residual garnet have high (La/Yb)\(_\text{N}\) and steep HREE slopes.
Memory Tricks
Don’t Confuse
| This | vs | That |
|---|---|---|
| Isotopes: same \( Z \) | Isobars: same \( A \) (the parent–daughter pairs of geochronology) | |
| Camouflage: no fractionation (Hf in zircon) | Capture: trace ion enriched early (Ba in K-feldspar) | |
| Compatible (\( D > 1 \)): Ni, Cr, Co, Sc | Incompatible (\( D < 1 \)): LILE, HFSE, LREE | |
| LILE: fluid-mobile, low \( Z/r \) | HFSE: fluid-immobile, high \( Z/r \) | |
| Positive Eu anomaly: plagioclase accumulation | Negative Eu anomaly: plagioclase removed (fractionated or residual) | |
| Lanthanide contraction: Zr ≈ Hf in radius | Zr/Hf still varies slightly; it is not perfectly constant |
Common Mistakes
Practice
All questions below are original practice questions written for this page. They are not from any CSIR NET paper. Use the radii in the table above.
Level 1: Foundation
- State Goldschmidt’s rule for the size limit on extensive ionic substitution.
- Which REE commonly occurs as a divalent ion in reduced magmas?
- What coordination does a radius ratio of 0.55 predict?
Answers
- Radii should differ by less than about 15%.
- Europium (Eu\(^{2+}\)).
- Octahedral (CN 6): 0.55 lies between 0.414 and 0.732.
Level 2: Part B Style
Practice Q1. Which trace element is most likely to be enriched in early-crystallising olivine from a basaltic melt?
(a) Rb (b) Ni (c) Zr (d) La
Answer: (b) Ni
\( \text{Ni}^{2+} \) (0.69 Å) differs from \( \text{Mg}^{2+} \) (0.72 Å) by about 4%, has the same charge and gains a large octahedral site preference energy (d\(^8\)). Rb, Zr and La are far too large or too highly charged for the olivine M site and are incompatible.
Practice Q2. Barium enters K-feldspar in preference to potassium during crystallisation of a granite. In Goldschmidt’s terms this is:
(a) camouflage (b) capture (c) admission (d) exsolution
Answer: (b) capture
\( \text{Ba}^{2+} \) (1.42 Å) is about 6% smaller than \( \text{K}^{+} \) (1.51 Å) and has a higher charge, so it is captured preferentially by early K-feldspar (charge balance by \( \text{Ba}^{2+}\text{Al}^{3+} \rightleftharpoons \text{K}^{+}\text{Si}^{4+} \)). Camouflage would require the same charge.
Practice Q3. Which element group is used to classify altered, metamorphosed basalts by tectonic setting because it is immobile in aqueous fluids?
(a) Rb, Cs, Ba (b) K, Na, Ca (c) Zr, Nb, Ti, Y (d) Sr, Pb, Eu\(^{2+}\)
Answer: (c) Zr, Nb, Ti, Y
These are high field strength elements (or, for Y, a HREE-like element). Their high \( Z/r \) makes them insoluble in aqueous fluids, so their ratios survive alteration. The other options are fluid-mobile LILE or major cations.
Practice Q4. A granite has chondrite-normalised Sm\(_\text{N}\) = 50, Eu\(_\text{N}\) = 24 and Gd\(_\text{N}\) = 32. Its Eu/Eu\(^*\) and the most likely interpretation are:
(a) 0.60; plagioclase fractionation (b) 0.60; plagioclase accumulation (c) 1.67; plagioclase accumulation (d) 0.48; garnet fractionation
Answer: (a) 0.60; plagioclase fractionation
\( \text{Eu}^* = \sqrt{50\times 32} = \sqrt{1600} = 40 \); \( \text{Eu}/\text{Eu}^* = 24/40 = 0.60 \). A value below 1 is a negative anomaly: Eu\(^{2+}\) was removed with plagioclase. Option (d) uses \( 24/50 \), not the geometric mean.
Level 3: Part C Style
Practice Q5. Using the Shannon radii in the table, consider these statements:
- \( \text{Sr}^{2+} \) can replace \( \text{Ca}^{2+} \) in plagioclase, because their radii at CN VIII differ by about 12.5%.
- \( \text{Li}^{+} \) replacing \( \text{Mg}^{2+} \) in a pyroxene is an example of admission, and needs a coupled substitution.
- The radius ratio of \( \text{Ca}^{2+} \) (CN VIII) to \( \text{O}^{2-} \) predicts octahedral coordination.
- \( \text{Hf}^{4+} \) follows \( \text{Zr}^{4+} \) so closely because of the lanthanide contraction.
Which statements are correct?
(a) 1, 2 and 4 only (b) 1 and 3 only (c) 2, 3 and 4 only (d) 1, 2, 3 and 4
Answer: (a) 1, 2 and 4 only
- 1 true: \( (1.26 - 1.12)/1.12 = 0.125 \), within the 15% limit; same charge (camouflage-like, with slight preference).
- 2 true: Li\(^{+}\) has a lower charge than Mg\(^{2+}\) (admission); neutrality needs, for example, \( \text{Li}^{+}\text{Fe}^{3+} \rightleftharpoons 2\text{Mg}^{2+} \) or \( \text{Li}^{+}\text{Al}^{3+} \).
- 3 false: \( 1.12/1.40 = 0.80 \), which is in the 0.732–1.0 range: cubic (CN 8).
- 4 true: the 14 lanthanides between La and Hf shrink the radius so that Hf\(^{4+}\) (0.83 Å) almost equals Zr\(^{4+}\) (0.84 Å).
Practice Q6. Two basalts from one volcanic province were analysed (ppm):
| Sample | Ba | Nb | Zr | Rb | Ni |
|---|---|---|---|---|---|
| A | 600 | 3 | 80 | 30 | 40 |
| B | 150 | 30 | 160 | 10 | 250 |
Which interpretation fits the data best?
(a) A is a primitive mantle melt and B an evolved arc lava (b) A carries a subduction (slab-fluid) signature; B is a relatively primitive within-plate basalt (c) Both formed from the same source; the differences reflect alteration of B (d) A is more primitive than B because its Ba is higher
Answer: (b)
- Ba/Nb: A = 600/3 = 200; B = 150/30 = 5. Very high Ba/Nb with very low Nb is the classic arc signature: fluid-mobile LILE added, HFSE (Nb) left behind in the slab.
- Ni: B has 250 ppm, A only 40 ppm. Ni is compatible in olivine, so B has lost less olivine and is more primitive. That rules out (a) and (d).
- Alteration (c) mobilises Ba and Rb, not Nb and Zr; it cannot create a tenfold Nb difference.
Quick Revision Table
| Concept | Key point |
|---|---|
| Isobars | same \( A \); parent–daughter pairs (\( ^{87}\text{Rb}/^{87}\text{Sr} \)) |
| Cosmic abundance | H, He dominant; Li–Be–B deficit; Fe peak; Oddo–Harkins saw-tooth |
| Ionic character | \( 1 - e^{-0.25\Delta\chi^2} \); Si–O about 45%, K–O about 82% |
| Radius ratio | 0.225 (4), 0.414 (6), 0.732 (8), 1.0 (12) |
| Ionic potential | below 3 soluble; 3–12 hydrolysates; above 12 oxyanions |
| Goldschmidt | 15% size, ±1 charge, smaller and higher-charge ions favoured |
| Ringwood | lower electronegativity favoured (Cu not in Na sites; Zn not in olivine) |
| Camouflage / capture / admission | Hf for Zr / Ba for K / Li for Mg, Rb for K |
| Crystal field | Ni, Cr strongly octahedral; Mn\(^{2+}\), Fe\(^{3+}\), Zn none |
| LILE vs HFSE | fluid-mobile K, Rb, Cs, Sr, Ba, Pb vs immobile Zr, Hf, Nb, Ta, Ti, Th |
| Eu/Eu\(^*\) | \( \text{Eu}_\text{N}/\sqrt{\text{Sm}_\text{N}\text{Gd}_\text{N}} \) |
Key Takeaways
What’s Next
Ionic substitution makes solid solutions; to say how stable they are, and how many phases can coexist, you need thermodynamics. Continue with Phase Rule and Thermodynamics of Solutions and Fluids. For the quantitative version of compatibility, see Distribution Coefficients, Nucleation and Diffusion.