No joint field · Bridge A = SPECULATIVE pedagogy

Labels: SOURCED HYPOTHESIS SPECULATIVE OPEN NONE (REAL = same-domain materials or linguistics, shown like SOURCED)

Path E · Ranking-edge chemistry scout

Spinel dopant site preference and spec point 6 — chemistry depth note,

What’s claimed and how well supported

Labels on this page
SPECULATIVE ×3
Strongest label used
SPECULATIVE
Main caveat (quoted)
The comparison to Navajo is a ◆ SPECULATIVE test against a specification, not a correlate, and not a joint field.

Counted automatically from the labels printed on this page. Counts are a reading aid, not a grade of the page.

Terms on this page (4), plain-language definitions
Host (matrix)
Majority element/phase whose lattice dominates the material.
Dopant
Intentional dilute impurity meant to tune electronic/optical/catalytic behavior without replacing the host as majority phase.
Solid solution
Solute atoms dissolved into the host lattice (substitutional or interstitial) without a separate precipitate phase.
Valence (chem.)
Bonding capacity / oxidation-related electron count of an atom in context.

Definitions quoted from the Chem for linguists primer glossary.

Date: 2026-10-05 PT (weekday evening research push)
Status: Chemistry depth note. The chemistry is same-domain description from the sources opened below (REAL inside solid-state chemistry and mineralogy). The comparison to Navajo is a SPECULATIVE test against a specification, not a correlate, and not a joint field.
Audience: Grant (linguistics; less chemistry).
Does not: claim that any cation is a chemical fourth person; supply a stand-in for a theta role; invent numbers or site occupancies; revise Bridge A; move spinel to the molecule level, where it does not sit.

Why this note. The midday fourth-person note (Fourth person vs ranking §6) added optional spec point 6: a way to take one participant out of the ranked set, after which the default alignment survives the clash. Its open question 4 asked for the spinel candidate to be tested against point 6. The spinel scout (Spinel inversion ranking edge §6) left open question 1 (which "second scale": crystal field or Burdett's orbital size) and open question 4 (dopantIntentional dilute impurity meant to tune electronic/optical/catalytic behavior without replacing the host as majority phase. site preference in spinel hostsMajority element/phase whose lattice dominates the material., to tie the case to the doping strand). This note takes up all three.

What is new (one sentence). Spinel chemistry does have a sourced category that the d-electron scale does not apply to — d⁰, high-spin d⁵ and d¹⁰ ions, zero crystal-field site preference — but taking a cation into that category does not make the charge default survive: other scales (cation size, sp³ covalency, electrostatics) still rank it, so Zn²⁺, Mn²⁺ and Cd²⁺ ferrites stay normal while MgFe₂O₄ and MgGa₂O₄ stay inverse, and Price, Price & Burdett (1982) sort 98% of 172 spinels without counting d electrons at all — so point 6 at most partly passes: the exit category exists, but what point 6 says should follow from it does not.


1. Sources opened this run

Source How read Used for
O'Neill & Navrotsky 1983, "Simple spinels: crystallographic parameters, cation radii, lattice energies, and cation distribution," American Mineralogist 68: 181–194 (MSA open archive PDF) Page images, printed pp. 181, 183–185, 188 (rendered with pdftoppm; text layer two-column garbled, used only to locate) Zero-CFSE ions: crystal-field model "has nothing at all to say"; CFSE model wrong in about half of applicable cases (their citation of Glidewell 1976); Zn sp³ as ad hoc; size rule; Table 1 normal/inverse codes; Ca²⁺ in magnetite (their citation of de Sitter et al. 1977)
Sanchez-Lievanos, Stair & Knowles 2021, "Cation Distribution in Spinel Ferrite Nanocrystals…," Inorganic Chemistry 60: 4291–4305 (NSF PAR accepted copy) PDF text layer, pp. 4291–4292, 4299–4301 incl. Table 2 Three enthalpic factors (size, electrostatics, CFSE) plus entropy; bulk inversion parameters for six ferrites; Zn 4s/4p tetrahedral preference; Ni²⁺ exclusively octahedral across Ni₁₋ᵧZnᵧFe₂O₄ (their report of neutron diffraction, their ref. 94)
Price, Price & Burdett 1982, "The factors influencing cation site-preferences in spinels: a new Mendelyevian approach," Physics and Chemistry of Minerals 8: 69–76 Abstract only (Springer page) Orbital-radius index r_σ = r_s + r_p; 172 chalcogenide spinels; 98% sorted; no d-electron count needed
Bosi, Hålenius, D'Ippolito & Andreozzi 2012, "Blue spinel crystals in the MgAl₂O₄–CoAl₂O₄ series: Part II…," American Mineralogist 97: 1834ff. Abstract only (Sapienza IRIS record) Co²⁺ (d⁷) marked tetrahedral preference over Mg in MgAl₂O₄; inversion 0.24 → 0.13 with Co; "steric factors"
Khaidukov et al. 2021, "Time- and Temperature-Dependent Luminescence of Manganese Ions in Ceramic Magnesium Aluminum Spinels," Materials 14: 420 (MDPI open access) PDF text layer, pp. 2–3, 10–11 Mn²⁺ (d⁵) on the tetrahedral site in MgAl₂O₄, argued from ionic radii; Mn⁴⁺ (d³) and Mn³⁺ (d⁴) on octahedral sites
LibreTexts, UC Davis Chem 124A §9.10 "Octahedral vs. Tetrahedral Geometries" (Mike Morris, 2001) Web text CFSE table (Δ_tet = 4/9 Δ_o): difference 0 for d⁰, d⁵, d¹⁰; 0.84 Δ_o for d³, d⁸; ordering d³,d⁸ > d⁴,d⁹ > d²,d⁷ > d¹,d⁶ > d⁰,d⁵,d¹⁰. This is the molecular-complex form of the same table the spinel scout used
Porta, Stone & Turner 1974, "The distribution of nickel ions among octahedral and tetrahedral sites in NiAl₂O₄–MgAl₂O₄ solid solutionsSolute atoms dissolved into the host lattice (substitutional or interstitial) without a separate precipitate phase.," J. Solid State Chem. 11: 135–147 Search-engine abstract extract only (ScienceDirect page 403 to both fetch and curl) Ni²⁺ predominantly octahedral for x = 0.01–0.55; labeled as extract wherever used

Not reached: Burdett, Price & Price 1982 JACS 104: 92–95 (ACS returned a bot-check page to curl and timed out to fetch; second failure, not retried further). Lombard et al. 2009 (HAL, Mn²⁺/Ti³⁺ EPR in MgAl₂O₄) and Dugué et al. (HAL, Ni:MgAl₂O₄ glass-ceramics) — HAL returned bot-check pages. Not opened: de Sitter et al. 1977 (Ca-magnetite; used only via O'Neill & Navrotsky's sentence, labeled), Glidewell 1976, Navrotsky & Kleppa 1967, Dunitz & Orgel 1957, McClure 1957.


2. Sourced chemistry

2.1 (a) The category the d-electron scale does not apply to

The category is real and sourced. In the crystal-field account the octahedral site preference is the CFSE difference between octahedral and tetrahedral coordination. LibreTexts' table gives that difference as 0 for d⁰, high-spin d⁵ and d¹⁰. Common spinel cations in this set: Mg²⁺ and Al³⁺ (no d electrons), Mn²⁺ and Fe³⁺ (high-spin d⁵), Zn²⁺, Ga³⁺, Cd²⁺, In³⁺ (d¹⁰).

The crystal-field scale is silent on these ions, by its own lights. O'Neill & Navrotsky (p. 183): "many of the ions that occur commonly in the spinel structure do not have any CFSE, and therefore, in cases where both ions are of this type, the model has nothing at all to say about which ion prefers which site." They add that even where it applies, "in about half of the cases … it does not predict the observed cation distribution correctly" (their citation of Glidewell 1976, not opened).

But the ions do not then fall back to the charge default. Each zero-CFSE pair still has a definite arrangement, and it is not always normal:

Spinel (both ions zero CFSE) Arrangement Source
MgAl₂O₄ normal (x < 2/3); x rises with T O'Neill & Navrotsky Table 1; spinel scout §2
ZnGa₂O₄ normal O'Neill & Navrotsky Table 1
ZnFe₂O₄ normal, x = 0 Table 1; Sanchez-Lievanos Table 2
MnFe₂O₄ normal, x = 0.20 O'Neill & Navrotsky p. 188 ("indeed normal"); Sanchez-Lievanos Table 2
CdFe₂O₄ normal O'Neill & Navrotsky p. 188
MgFe₂O₄ inverse, x = 0.90 Table 1 (I); Sanchez-Lievanos Table 2
MgGa₂O₄ inverse Table 1 (I)
MgIn₂O₄ the only case where electrostatic energy favors the inverse O'Neill & Navrotsky p. 188, point (5)

Flag: O'Neill & Navrotsky's Table 1 prints "MgFe₂O₄" twice, once N (u = 0.2615) and once I (u = 0.257). The I row matches the 2021 review (x = 0.90); the N row is probably a misprint (MnFe₂O₄ would fit), but this note does not correct the source and uses only the I row.

What decides instead (as the sources state it):

  • Size. O'Neill & Navrotsky (abstract; p. 188): "the size of a cation is important in determining its site preference"; in 2–3 spinels electrostatics tend to put the larger ion on the tetrahedral site (series MgB₂O₄ with B = Al < Ga < Fe < In; AFe₂O₄ with A = Ni < Mg ≈ Zn < Fe²⁺ < Mn < Cd). The 2021 review states the size factor the other way round ("generally favors placing the smaller cation in the smaller tetrahedral sites"). The two sources disagree on the direction of the size effect; this note records both and picks neither.
  • Covalency (Zn²⁺). The review: "the valenceBonding capacity / oxidation-related electron count of an atom in context. 4s and 4p orbitals of Zn²⁺ prefer a tetrahedral bonding geometry because of their tendency to form strong sp³-type covalent bonds." O'Neill & Navrotsky (p. 183) call the zinc explanation an ad hoc extension that "could not be extended to many other cations," including Mg and Al in MgAl₂O₄.
  • Electrostatics and charge. The review: electrostatics favor "placing the cation with the larger charge in the octahedral sites" — this is the charge default — but which way it points can depend on the oxygen parameter u.
  • Entropy. Favors the random distribution (x = 2/3), significant only at high temperature (review; O'Neill & Navrotsky).

So a zero-CFSE ion exits one scale, not the ranking. Zn²⁺ keeps the normal arrangement because it carries its own positive tetrahedral preference (sp³ covalency on one account, size on another), not because it stands outside comparison. Mg²⁺ in a ferrite or gallate exits the d-scale and the arrangement is still inverse.

2.2 (b) Dopant and minor-cation site preference in spinel hosts

Sourced cases only. "Default" = the charge default (2+ on the tetrahedral site, 3+ on the octahedral).

Cation (d count) Host Site What the source says decides it Source (how read)
Ni²⁺ (d⁸, largest crystal-field octahedral preference) Ni₁₋ᵧZnᵧFe₂O₄ nanocrystals, y = 0 → 1 Octahedral only, "regardless of the concentration of Zn²⁺" Crystal-field preference of d⁸ (review's reading); "preference of Ni²⁺ for octahedral sites is stronger than the preference of Zn²⁺ for tetrahedral sites" Sanchez-Lievanos 2021, reporting neutron diffraction (opened)
Ni²⁺ (d⁸) NiₓMg₁₋ₓAl₂O₄, x = 0.01–0.55 Predominantly octahedral; octahedral fraction higher in the solid solutions than in NiAl₂O₄ (78% inverse at 1273 K); "linked with an increase in Dq"; in Ni-dilute solutions "Madelung energy may then be more effective in counteracting the octahedral stabilization" Crystal field vs electrostatics Porta, Stone & Turner 1974 (abstract via search extract only)
Co²⁺ (d⁷, small crystal-field octahedral preference) (Mg,Co)Al₂O₄, Co = 0.07–1.00 Tetrahedral, "marked preference of Co²⁺ for tetrahedral coordination with respect to Mg" Authors stress "the importance of steric factors" Bosi et al. 2012 (abstract opened)
Co²⁺ (d⁷) CoFe₂O₄ (host cation, not dopant) Largely octahedral, x = 0.93 — Sanchez-Lievanos Table 2
Mn²⁺ (high-spin d⁵, zero CFSE) MgAl₂O₄ Tetrahedral (green 520–525 nm emission); octahedral Mn²⁺ judged "highly unlikely" Ionic radius (Mn²⁺ 0.83 Å vs Al³⁺ 0.535 Å octahedral; Mg²⁺/Mn²⁺ 0.57/0.66 Å tetrahedral) Khaidukov et al. 2021 (opened)
Mn⁴⁺ (d³), Mn³⁺ (d⁴) MgAl₂O₄ Octahedral (substituting Al³⁺); Mn⁴⁺ needs charge compensation by extra octahedral Mg²⁺ Charge, ionic radius similar to Al³⁺ Khaidukov et al. 2021 (opened)
Ca²⁺ (d⁰) Ca-substituted magnetite All tetrahedral (Mössbauer) — "the most startling and emphatic illustration" of the larger-ion-to-tetrahedral tendency, though Ca²⁺ is "normally found in eight-fold coordination" Size de Sitter et al. 1977 as cited by O'Neill & Navrotsky p. 188 (de Sitter not opened)
Zn²⁺ (d¹⁰) Ni₁₋ᵧZnᵧFe₂O₄ nanocrystals Both sites; tetrahedral fraction rises with y and with annealing at 600 °C sp³ / 4s4p preference; nanocrystal kinetics Sanchez-Lievanos 2021 (opened)

Reading the table plainly.

  • One clean d-count flip at dopant level: Ni²⁺. A 2+ cation, which the charge default would put on the tetrahedral site, goes octahedral, and the sources tie that to its d⁸ crystal-field preference. This is the doping-strand version of spec point 3 (the second scale overrides the default).
  • Co²⁺ shows the d-scale losing. d⁷ has a small octahedral crystal-field preference (LibreTexts: 0.27 Δ_o), yet in MgAl₂O₄ Co²⁺ prefers the tetrahedral site even more than Mg²⁺ does; the authors credit steric factors. In the ferrite host CoFe₂O₄ the same ion is largely octahedral. Same ion, different slot by host.
  • Mn shows the confound. Mn²⁺ (d⁵) tetrahedral, Mn³⁺/Mn⁴⁺ (d⁴/d³) octahedral: the slot follows the oxidation state, but oxidation state changes the charge and the radius as well as the d count, and the source argues from radius. This is not a clean d-count test.
  • Zero-CFSE dopants follow size, not the default. Mn²⁺ goes where the charge default would put it (tetrahedral, on the Mg site) and Ca²⁺ goes tetrahedral in magnetite, both argued from size. Neither is "outside ranking."

2.3 (c) Which "second scale"

  • Burdett, Price & Price 1982 (JACS): not reached. Only the companion paper's abstract was opened.
  • Price, Price & Burdett 1982 (PCM, abstract): an orbital-radius index r_σ = r_s + r_p sorts 172 chalcogenide spinels into normal and inverse "with only four errors (a predictive success rate of 98%)"; "the gross sorting is achieved without recourse either to the number of d-electrons or an orbital radius r_d"; so "the energies and extents of the cation s and p-orbitals … primarily determine coordination number"; crystal-field ideas are "less generally applicable."
  • O'Neill & Navrotsky 1983: size is important; the crystal-field model is silent on many ions and wrong in about half the cases where it applies (via Glidewell); the change in electrostatic energy on disordering "is very large and would swamp any CFSE term," and near 2000 K is "about two orders of magnitude larger" than the configurational-entropy term (p. 188, point (1)) — this is their model calculation, not a measurement; Fe₃O₄ may need extra non-configurational entropy; MgAl₂O₄ "cannot be described adequately by any simple model."
  • Sanchez-Lievanos 2021: three enthalpic factors (size, electrostatics, crystal field) plus entropy; crystal field "and other bonding considerations" decide in some cases (NiFe₂O₄, ZnFe₂O₄).

What the opened sources support: the d-electron scale is not the second scale. It is one of several, and on Price, Price & Burdett's account not the primary sorter. Where it is strong (d³, d⁸: Cr³⁺, Ni²⁺) it shows clearly; elsewhere size, covalency and electrostatics carry the decision. The JACS paper itself would be needed to say more about Burdett's specific claim that crystal-field effects are a "small perturbation" (that wording comes only from the spinel scout's Wikipedia summary).


3. Against the spec (points 1–6)

Points 1–5 from Inverse ranking §6; point 6 from Fourth person vs ranking §6. Spinel column updated from the spinel scout with this run's sources.

Spec point Navajo (as sourced in the inverse and fourth-person notes) Spinel, with dopants Verdict
1. One unchanged event; two participants with fixed roles Same event; agent and patient fixed Same composition; charges fixed. Still a standing structure, no event, no entailed roles Partial (unchanged)
2. A second ranking scale independent of the roles Animacy (with definiteness, topicality) Independent of charge, yes; but plural — crystal field, size, sp³ covalency, orbital radius. Price, Price & Burdett sort 98% without d electrons Pass, weakened: there is no single second scale; the scales are several and the sources weigh them differently
3. Scales disagree → marked alignment preferred (62b) inverse good, (62a) direct ?* Ni²⁺ (d⁸) goes octahedral as host cation (NiFe₂O₄, NiAl₂O₄ inverse) and as a dilute dopant (MgAl₂O₄ at x = 0.01, extract; Ni–Zn ferrites, review). Co²⁺ shows the d-scale losing to steric factors Pass for strong d-scale ions; host-dependent for weak ones
4. Some alignments banned outright (56) , (60b) , one bi- per clause Still only the counting limit. Mn²⁺ on the octahedral site in MgAl₂O₄ is called "highly unlikely" on size grounds, but that is a strong preference, not a ban Partial (unchanged)
5. Both participants stay inside the unit Both remain pronominal arguments Dopants and host cations stay in the lattice; Mn⁴⁺ is charge-compensated inside the lattice, nothing leaves Pass
6. A way to take one participant out of the ranked set, so the default survives the clash Fourth person: "not involved in the animacy hierarchy"; with a low agent and a human patient, the fourth-person patient keeps the direct form (Willie (22) vs (20b) *) Exit category exists for the d-scale (d⁰, d⁵, d¹⁰; zero CFSE; crystal-field model "has nothing at all to say"). Default survival does not follow: MgFe₂O₄ (x = 0.90) and MgGa₂O₄ stay inverse; Zn²⁺/Cd²⁺/Mn²⁺ ferrites are normal because size or covalency rank them, not because they are unranked. On Price/Price/Burdett's orbital-radius scale every cation has a value, so there is no exit category at all. No bans tied to the inverse exist to be lifted Partial at most, leaning fail. The category exists; what point 6 says should follow from it does not

One asymmetry worth recording (analysis of the sourced cases, not a source's claim). In the inverse ferrites the zero-CFSE ion is the one displaced into the marked slot: in Fe₃O₄, NiFe₂O₄ and CoFe₂O₄ it is Fe³⁺ (d⁵) that goes tetrahedral. In Navajo the participant taken out of the ranked set (fourth person) is the one whose presence lets the default alignment stand. So in spinel, "outside the d-scale" correlates with moving under the marked arrangement; in Navajo, "outside the hierarchy" correlates with not triggering it. The two do not line up.


4. Separation

SOURCED (do not inflate)

  • Zero crystal-field site preference for d⁰, high-spin d⁵, d¹⁰ (LibreTexts §9.10); the crystal-field model is silent on such pairs (O'Neill & Navrotsky p. 183).
  • Zero-CFSE pairs are not all normal: MgAl₂O₄, ZnGa₂O₄, ZnFe₂O₄, MnFe₂O₄, CdFe₂O₄ normal; MgFe₂O₄ (x = 0.90), MgGa₂O₄ inverse (Table 1; review Table 2; p. 188).
  • Size, covalency (Zn sp³) and electrostatics as the other site-deciding factors; the direction of the size effect differs between O'Neill & Navrotsky and the 2021 review.
  • Ni²⁺ octahedral across Ni–Zn ferrites (review); Co²⁺ tetrahedral in MgAl₂O₄ with steric factors (Bosi et al. abstract); Mn²⁺ tetrahedral and Mn³⁺/Mn⁴⁺ octahedral in MgAl₂O₄ (Khaidukov et al.); Ca²⁺ tetrahedral in magnetite (de Sitter via O'Neill & Navrotsky, secondary).
  • Price, Price & Burdett: 98% of 172 spinels sorted by r_σ without d-electron count (abstract).
  • Ni²⁺ predominantly octahedral in dilute NiₓMg₁₋ₓAl₂O₄ — abstract via search extract only.

HIS SPECULATIVE (not findings)

  • Grant's argument-ranking frame and the six-point spec as his shape for a cross-domain edge. Not tested as a joint field; tested only as a checklist.
  • Phrase ↔ molecule / word ↔ atom as the Zipf-shared level (Grant). Not tested here; spinel and its dopants are at the crystal level, one cut off his molecule level.

Not claimed

  • That a zero-CFSE ion, a dopant, or any site is a chemical fourth person, a theta role, or an obviative. No source says so; the comparison in §3 shows they behave differently.
  • That the crystal-field scale is the right "second scale" or the wrong one; the opened sources disagree on weight, and the JACS paper was not reached.
  • Any site occupancy, inversion value or energy not printed in a source opened this run.
  • Bridge A is not revised.

5. Verdict

Point 6: partial at most, leaning fail. Spinel chemistry has a sourced exit category for its d-electron scale (d⁰, d⁵, d¹⁰), so the first half of point 6 is there. The second half — that the default then survives — is not: those ions are still ranked by size, covalency and electrostatics, and Mg²⁺ spinels with Fe³⁺ or Ga³⁺ stay inverse. On Price, Price & Burdett's orbital-radius scale there is no exit category at all. The Navajo fourth person is outside the hierarchy; a zero-CFSE ion is outside one of several.

Doping tie-in (spinel OQ4): answered with sources. The cleanest dopant case is Ni²⁺ (d⁸), which takes the octahedral site against the charge default as a host cation and in dilute solid solution and Ni–Zn ferrites. Co²⁺ and Mn²⁺ show the d-scale losing to or giving way to size.

Second scale (spinel OQ1): partly answered. The opened sources make the d-electron scale one of several, and not the primary sorter on Price, Price & Burdett's account; the JACS paper itself was not reached.

Overall standing of the spinel candidate: unchanged at the core (points 3 and 5 pass, 1 and 4 partial), slightly weaker on point 2 (no single second scale), partial-to-fail on point 6. Still a candidate, not a correlate, and still at the crystal level. The search for a molecule-level case (spinel OQ2) continues.


6. Open questions

  1. Open Burdett, Price & Price 1982 JACS itself (two access failures this run) for the "small perturbation" claim and their treatment of Al³⁺ and Zn²⁺.
  2. Open Porta, Stone & Turner 1974 beyond the abstract extract (dilute-Ni regime; their "tentative evidence that α may increase" below x = 0.25 — whether the d-scale weakens as Ni becomes a true trace dopant).
  3. Is there any spinel system where removing the d-scale (substituting a zero-CFSE ion) restores the normal arrangement and nothing else (size, covalency) can explain it? That would be the sharpest point-6 test. None found in the sources opened.
  4. Molecule-level analogue (spinel scout OQ2) — still the main open chemistry question for Grant's Zipf level; point 6 should be part of that test too (e.g., whether a d⁰/d¹⁰ metal center in a mixed-ligand complex drops out of a site-preference ranking).
  5. Grant's multiple-wh test remains with his texts; nothing here bears on it.

Pointer (2026-10-05 late evening): open question 4 is taken up in TBP apicophilicity ranking edge (molecule level: groups for which apicophilicity "cannot be rigorously defined" either break the bipyramid or keep only the apical slot, so point 6 again leans fail).

Corpus note: spinel-dopant-site-preference-point6-v0.md. The chemistry is same-domain. The comparison to Navajo is speculative and is not a correlate. Not Bridge A. Not a joint field.