Depth note — Electronic doping Family 1 contrast: Si · GaN:Mg · oxide
What’s claimed and how well supported
- Labels on this page
- SPECULATIVE ×9
- Strongest label used
- SPECULATIVE
- Main caveat (quoted)
- Materials first; Bridge A only after.
Counted automatically from the labels printed on this page. Counts are a reading aid, not a grade of the page.
Terms on this page (7), 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.
- Vacancy
- Point defect: a missing atom site in the lattice.
- Solute
- Intentional minority species dissolved in the host (alloying addition).
- Valence (chem.)
- Bonding capacity / oxidation-related electron count of an atom in context.
- Dislocation
- Line defect enabling plastic slip at realistic stresses; edge vs screw are the two classic geometries.
- HCP
- Hexagonal close-packed — fewer primary slip systems; Mg, Ti, Zn often anisotropic.
Definitions quoted from the Chem for linguists primer glossary.
Date: 2026-09-21 PT
Audience: syntax/semantics readers building alloy-design literacy (Chem for linguists)
Purpose: One linguist-readable contrast of the three TRUE electronic-doping hostsMajority element/phase whose lattice dominates the material. already in the corpus (mechanism taxonomy Family 1) — shared carrier-tuning surface, divergent ionization / maturity / “over” physics, and vocabulary that stays dopantIntentional dilute impurity meant to tune electronic/optical/catalytic behavior without replacing the host as majority phase. / donor / acceptor. Materials first; Bridge A only after. Prefer depth/synthesis over new hosts.
Locks: Bridge A locked as SPECULATIVE supporting case only. Materials facts first; linguistic parallels only in a flagged box. No joint-field claim. Dowty / Triskelion / Merge·Agree / kanji / Navajo θ-typing not advanced here.
Hard separations: shallow Si ≠ deep GaN:Mg ≠ oxide mid-gap / V_O; electronic doping ≠ intercalation (Family 2) ≠ NMC “papers say doping” ≠ Al–Mg SSS; continuous N_D/N_A ≠ discrete arity.
Companion stack: Mechanism taxonomy §1 (Family 1); Si:P/As/B; GaN:Mg; Oxide electronic; Physics of doping; Doping vs alloying vocabulary (Path D); CNT / graphene (Family 2 foil); NMC:Mg cathode (NMC foil); Al–Mg (SSS foil); Reading order §2; Conclusions.
- 0. One-screen claim
- 1. Shared Family 1 surface (REAL — read once)
- 2. Si:P/As/B — textbook shallow; industrial CMOS baseline
- 3. GaN:Mg — industrial deep acceptor; H activation co-actor
- 4. Oxide (TiO₂ / ZnO-class) — mid-gap / V_O; research ↔ TCO
- 5. Side-by-side (Family 1 contrast)
- 6. Hard splits — Family 1 ≠ foils
- 7. Vocabulary lock (Path D — name once)
- 8. SPECULATIVE Bridge A after materials
- 9. What to do next (reader checklist)
- 10. Sources / corpus pointers
0. One-screen claim
Three corpus hosts share intentional substitutional (etc.) foreigners that tune carrier type and density — not yield strength — without sharing one ionization depth, one maturity, or one “over” physics:
| Peg | Host lattice | Foreigner | Shallow vs deep / maturity | Headline meter | Typical “over” |
|---|---|---|---|---|---|
| Si:P/As/B | Diamond-cubic Si (E_g ~1.12 eV) | P, As (donors); B (acceptor) | Shallow (~45–54 meV); fully industrial CMOS / IC | Net N_D−N_A; extrinsic n/p | Impurity band / Mott; clustering; μ collapse; compensation |
| GaN:Mg | Wurtzite GaN (E_g ~3.4 eV) | Mg_Ga acceptor (+ H passivation / activation) | Deep (~135–200 meV class); industrial LED/laser | SIMS [Mg] vs Hall free p after N₂ anneal | Self-compensation; PIDs / clusters; μ / optical loss; failed H activation |
| Oxide (TiO₂ / ZnO-class) | Wide-gap oxide (+ often V_O) | Subst. metal (Fe/Cu/Nb on Ti; Al on Zn; …) | Mid-gap / V_O—not H-like shallow; research-leaning (photo) / semi-industrial (AZO TCO) | Visible QE / mid-gap (photo); nμ / ρ (TCO) | Recombination centers; μ collapse; secondary oxides; vacancyPoint defect: a missing atom site in the lattice. compensation |
Pedagogy in one line: Family 1 is a shared mechanism name (carriers / bands / Fermi level) — not one concentration number or one failure mode. Name ionization depth + process clock before hanging under/useful/over.
Vocabulary lock (all three): prefer dopant / donor / acceptor. Do not call percent-level metal solutesIntentional minority species dissolved in the host (alloying addition). (Al–Mg SSS, Mn in TWIP, …) “dopants.” Reserve dopant for dilute electronic (etc.) intent — these three are the corpus’s cleanest yes-cases.
1. Shared Family 1 surface (REAL — read once)
What these three share — and what they do not.
1.1 Shared
| Theme | What it means |
|---|---|
| Intentional foreigners | Dilute impurities placed to change electrical / optoelectronic behavior |
| Substitutional (primary) | Foreign atom on a host lattice site (Si site; Ga site; Ti/Zn cation site) — oxide also co-engineers vacancies |
| Carrier type / density | Success meter is n or p (or mid-gap absorption / charge separation), not yield strength or creep |
| Honest word | Dopant — materials community’s default meaning (Mechanism taxonomy §1; Path D) |
1.2 Not shared
| Not this | Why |
|---|---|
| One ionization energy | Si ~45 meV ≠ GaN:Mg ~135–200 meV ≠ oxide mid-gap / V_O |
| One useful cm⁻³ or at.% | Si device windows often ~10¹⁵–10¹⁸ cm⁻³; GaN [Mg] ~10¹⁹ with p ~10¹⁷; oxide photo/TCO often at.% class |
| One maturity | Fully standardized CMOS ≠ industrial GaN LED fab ≠ research photocatalysis campaigns |
| One “over” physics | Mott MIT ≠ self-compensation / PIDs ≠ recombination-center overrun |
| Permission to call Al–Mg SSS “doped” | Different mechanism family (Family 3) |
Design rule (unchanged): diagram → name mechanism = Family 1 → name failure past the window → which process/activation clock → only then SPECULATIVE Bridge A count mnemonic.
2. Si:P/As/B — textbook shallow; industrial CMOS baseline
Pointer: Si:P/As/B. Companion: Physics of doping §1–2.
What it is (REAL)
Crystalline diamond-cubic Si with substitutional group-V donors (P, As; also Sb) and group-III acceptors (B). Intrinsic n_i ~10¹⁰ cm⁻³ at 300 K is too low / untyped for transistors — doping sets majority-carrier type and density. Hydrogenic shallow levels (~45–54 meV) → ~fully ionized at 300 K in the dilute extrinsic window. This is the maturity ceiling Grant’s takeaway contrasts with research oxide doping, and the reference chassis against which GaN:Mg depth and oxide mid-gap stories are measured.
What is “saturated” / useful
Extrinsic plateau at device T: enough ionized net doping |N_D − N_A| for the intended n or p, typically still non-degenerate in channel/well/epi class (often ~10¹⁵–10¹⁸ cm⁻³ order — process-dependent; see checklist). Source/drain / contact doping may intentionally sit in the heavy / degenerate ~10¹⁹+ class. “Saturated” ≠ max implant dose — it means the useful extrinsic window for that device region.
What “over” looks like (plural)
| Overrun mode | What goes wrong |
|---|---|
| Impurity band / Mott MIT | Discrete H-like levels merge near N_crit ~few×10¹⁸ cm⁻³ class (Si:P ~3.7×10¹⁸; Si:As ~8.5–8.6×10¹⁸; Si:B ~4×10¹⁸ order — checklist) |
| Degenerate statistics | E_F enters band; Fermi–Dirac required (~10¹⁹ class) |
| Clustering / precipitation | Inactive aggregates (B clusters; As clustering) |
| Mobility collapse | Ionized + neutral impurity scattering |
| Compensation | Opposite-type impurities cancel net carriers |
| Implant damage not annealed | Process failure — inactive dopants (not solubility “over”) |
| CZ growth instability | Constitutional supercooling ≳~10²⁰ cm⁻³ class |
Process / activation clock
| Clock | Si reading |
|---|---|
| Diagram / solubility | Equilibrium solubility vs implant metastable profiles |
| Process | Ion implant + RTA/anneal (CMOS default); CZ melt doping; diffusion; epi |
| Service | Later thermal budgets move dopants (TED for B); activation state is anneal-written |
Kinetic one-liner: implant places dose; anneal activates substitutional electrical activity and repairs damage; later budgets redistribute. Three clocks still apply (Process metastability).
Vocabulary
Dopant / donor / acceptor / doping — unequivocally correct. Never “alloying silicon” for CMOS carrier intent.
3. GaN:Mg — industrial deep acceptor; H activation co-actor
Pointer: GaN:Mg.
What it is (REAL)
Wurtzite GaN (wide gap ~3.4 eV) with Mg on Ga (Mg_Ga) as the primary p-type acceptor for LEDs/lasers. n-type GaN is comparatively easy; p-type was historically the bottleneck. Mg acceptor depth (~135–160 meV Hall class; some estimates ~190–200 meV) ≫ Si:B (~45 meV) → only a fraction of Mg atoms yield free holes at 300 K. MOCVD co-incorporates H → electrically neutral Mg–H complexes; as-grown ρ often ~10⁶ Ω·cm until N₂ activation anneal (typically ≳600–700 °C class) depassivates.
What is “saturated” / useful
Useful SIMS [Mg] often ~1–3×10¹⁹ cm⁻³ (peak Hall p often near ~2×10¹⁹ in classic series) after successful N₂ activation → free p typically peak ~4–6×10¹⁷ cm⁻³ (≪ [Mg] because of incomplete ionization + residual compensation). “Saturated” is peak free p for the device stack, not max [Mg].
What “over” looks like (plural)
| Overrun mode | What goes wrong |
|---|---|
| Incomplete ionization (baseline, not “over”) | Deep acceptor → p ≪ [Mg] even inside the useful window |
| H passivation | Neutral Mg–H — as-grown or failed / skipped anneal |
| Self-compensation | p falls while SIMS Mg rises past ~2–3×10¹⁹ (V_N / Mg_I / complexes) |
| PIDs / clustering | Pyramidal inversion domains; Mg-rich clusters near ~10²⁰ |
| μ / optical loss | Scattering; defect luminescence; laser-cavity pain |
| Over-anneal stack damage | InGaN QW In segregation; surface V_N; Mg diffusion |
Honest split: incomplete ionization is baseline physics inside the useful window; self-compensation / PIDs are overrun past it. Do not conflate them.
Process / activation clock (first-class for this host)
| Clock | GaN:Mg reading |
|---|---|
| Diagram / defect chemistry | Mg_Ga vs V_N / Mg_I compensation landscape |
| Process | MOCVD H incorporation → Mg–H; N₂ RTA activates; atmosphere can re-passivate |
| Service / stack | Later H-rich steps; thermal budget vs InGaN QW integrity |
Kinetic one-liner: growth captures passivated Mg–H; anneal writes free acceptors; wrong later ambient erases activation. Process atmosphere belongs on the doping page.
Vocabulary
Dopant / p-type doping / acceptor — correct. Same honesty as Si; not Al–Mg SSS “doped aluminum.”
4. Oxide (TiO₂ / ZnO-class) — mid-gap / V_O; research ↔ TCO
Pointer: Oxide electronic. Companion: Physics of doping §3.
What it is (REAL)
Wide-gap oxides with intentional substitutional metal dopants plus oxygen-vacancy (V_O) co-actors:
| Host | Goal | Dopant story (checklist order-of-magnitude) |
|---|---|---|
| TiO₂ (anatase/rutile) | Visible photocatalysis; charge separation | Fe, Cu, Nb, … — mid-gap / d-states + V_O / Ti³⁺; optima often ~0.5–few % (process-dependent) |
| ZnO (wurtzite; Al:ZnO / AZO) | Transparent conductor (low ρ + high visible T) | Al³⁺ on Zn²⁺ shallow-ish donor; useful Al often ~1–3 at.% class |
| Brief siblings | WO₃ photo; Ni-rich NMC cathode | V_O engineering; NMC is structural stabilizer — foil in §5, not merge |
Do not export Si shallow-donor formulas unchanged to TiO₂:Fe (Physics of doping open Q #6). Vacancies are lattice defects, not “just another dopant atom.”
What is “saturated” / useful
| Goal | Useful state |
|---|---|
| Photocatalysis | Dilute–few % where mid-gap absorption + temporary trapping improve net QE / degradation — without recombination-center overrun |
| AZO TCO | ~1–3 at.% Al (process-dependent) where product nμ minimizes ρ while visible transmittance stays high |
What “over” looks like (plural)
| Overrun mode | What goes wrong |
|---|---|
| Recombination-center overrun | Useful traps become e⁻/h⁺ meeting points (classic Fe:TiO₂ above ~0.5–1% class optima in cited lineages) |
| Vacancy compensation | Native acceptors (e.g. V_Zn) cancel intentional donors |
| Mobility collapse | Ionized-impurity + GB scattering (AZO Hall μ vs Al%) |
| Secondary oxides / segregation | Fe₂O₃ / CuO / Al₂O₃-like / ZnAlₓOᵧ phases |
| Deep-trap / redox-power loss | Excess V_O in WO₃-class → weak oxidizing power |
| Impurity-band / degeneracy | Electronic MIT family — keep separate from second-phase solvus overrun |
Process / activation clock (V_O co-actor)
| Clock | Oxide reading |
|---|---|
| Diagram / defect chemistry | Aliovalent dopant valenceBonding capacity / oxidation-related electron count of an atom in context. ↔ V_O formation energy (lower-valence often promotes V_O; higher may suppress) |
| Process | Synthesis p(O₂) / anneal / ALD–sol-gel capture non-equilibrium defects |
| Service | Vacuum-made V_O can relax in air; long-term stability / re-oxidation |
Kinetic one-liner: reducing anneal buys V_O / reduced centers; air / oxygen spends them. Same nominal at.% can be lattice donor, GB-inactive, or secondary-phase — characterization (XRD/XPS/EPR/Hall/PL) is non-negotiable.
Vocabulary
Dopant / doping — correct for carrier / mid-gap / TCO intent. Battery-cathode Mg/Al/Nb in NMC often papers say doping but meter ≠ n/p — see §5 foil.
5. Side-by-side (Family 1 contrast)
| Si:P/As/B | GaN:Mg | Oxide TiO₂/ZnO | |
|---|---|---|---|
| Ionization | Shallow (~45–54 meV); ~full at 300 K in dilute window | Deep (~135–200 meV); p ≪ [Mg] | Mid-gap / V_O — not H-like shallow Si |
| Useful meter | Net N_D−N_A; n ≈ net doping on extrinsic plateau | Hall free p after activation (≠ SIMS [Mg]) | Photo QE / mid-gap; TCO nμ |
| Typical useful window | Often ~10¹⁵–10¹⁸ cm⁻³ (device-region class) | [Mg] ~1–3×10¹⁹; p ~10¹⁷ class | at.% class (photo/TCO) |
| Process headline | Implant + RTA; CZ melt | MOCVD + H depassivation | p(O₂) / anneal; ALD/sol-gel |
| Maturity | Fully industrial SEMI / CMOS | Industrial LED/laser / GaN fabs | Photo often research; AZO semi-industrial |
| Headline “over” | Mott / clustering / μ | Self-comp / PIDs / failed activation | Recombination / μ / secondary oxide / V_O compensation |
| Honest word | Dopant | Dopant | Dopant |
| Bridge A electronic peg | Strongest textbook | Strongest industrial deep | Strongest research mid-gap / V_O |
Stacked pedagogy: Si trains the classical shallow extrinsic ladder; GaN:Mg trains that SIMS ≠ Hall p and that activation atmosphere is first-class; oxide trains mid-gap / vacancy co-actors and research maturity. Shared Family 1 label; plural physics.
6. Hard splits — Family 1 ≠ foils
6.1 Inside Family 1
Shallow Si ≠ deep GaN:Mg ≠ oxide mid-gap / V_O. Do not export Si ionization formulas to GaN:Mg or TiO₂:Fe. Do not treat “more dopant” as “more carriers” once self-compensation or recombination centers dominate.
6.2 Electronic doping ≠ intercalation (Family 2 foil)
| Family 1 (this note) | CNT / graphene intercalation | |
|---|---|---|
| Placement | Substitutional lattice site (+ oxide V_O) | Guest in channels / vdW gaps |
| Prefer saying | Dopant | Intercalant (dopant OK for carrier intent) |
| “Over” flavor | Mott / self-comp / recombination | Moisture dedoping; framework damage; loading past stability |
Pointer: CNT / graphene. Completes alloying vs doping vs intercalation triangle (Chem for linguists §5).
6.3 Electronic doping ≠ NMC “papers say doping”
| Family 1 | Ni-rich NMC:Mg/Al/Nb | |
|---|---|---|
| Headline meter | Carrier density / mid-gap / TCO ρ | Cycle life / thermal / cation order |
| Community verb | Dopant (true) | Papers often say “doping” |
| Mechanism honesty | Electronic | Structural / chemo-mechanical stabilizer |
Pointer: NMC:Mg cathode. Same-element Mg depth: Same-element Mg (GaN:Mg electronic ≠ NMC:Mg stabilizer ≠ Al–Mg solute).
6.4 Electronic doping ≠ Al–Mg SSS (Family 3 foil)
| Family 1 | Al–Mg AA5083 | |
|---|---|---|
| Slot filled | Carriers / bands | DislocationLine defect enabling plastic slip at realistic stresses; edge vs screw are the two classic geometries. pins (SSS) |
| Prefer saying | Dopant | Solute / alloying |
| “Over” | Mott / self-comp / recombination | GB β / sensitization |
| Never | Call Mg-in-Al a “dopant” | Call Si:P “solid-solution strengthening” |
Pointer: Al–Mg. Precipitation Family 4 contrast stays separate: Precipitation family §6 (precipitation ≠ carrier saturation).
7. Vocabulary lock (Path D — name once)
| Context | Prefer | Avoid |
|---|---|---|
| Si:P/As/B, GaN:Mg, TiO₂:Fe, AZO | Dopant / donor / acceptor | “Alloying silicon for CMOS carriers” as primary frame |
| Al–Mg, Cu–Ni, TWIP Mn, Cantor, CoCrMo, Zr SPPs | Solute / alloying / stabilizer / SPP | Calling percent-level metal solutes dopants |
| CNT intercalation | Intercalant (first); dopant OK for CT intent | Collapsing into SSS or oxide V_O talk |
| NMC:Mg | Stabilizer / papers say doping — name the meter | Equating with Si:P / GaN:Mg |
Full Path D: Doping vs alloying vocabulary.
SPECULATIVE Cross-domain section. Open after the materials.
8. SPECULATIVE Bridge A after materials
Everything in this section is pedagogical analogy. It does not make doping × θ-roles a joint field.
| Materials fact | SPECULATIVE Bridge A reading |
|---|---|
| Useful extrinsic Si window; peak GaN p after anneal; oxide photo/TCO sweet spot | Successful “saturation” of a named carrier configuration |
| Under-doped / passivated / UV-only | Under-filling relative to that configuration |
| Mott MIT; self-compensation / PIDs; recombination / secondary oxide | Plural “over” modes — count ladder is a memory peg, not one physics |
| Continuous N_D, N_A, at.%, Hall p | Continuous concentration ≠ discrete linguistic arity — do not force integer slots onto Mott curves or Kaufmann p(N_A) peaks |
| Strongest electronic Bridge A pegs in corpus | Si (textbook classical) · GaN:Mg (industrial deep + activation) · oxide (research mid-gap / V_O) — hang after naming Family 1 + which host |
| Family 1 ≠ intercalation ≠ NMC ≠ Al–Mg SSS | Hang Bridge A only after naming the mechanism family; never equate Ostwald ripening or GB β with Mott MIT |
Earn: one depth note that forces the reader to say which electronic host (shallow / deep / mid-gap) and which “over” before using under/useful/over; pairs the three true-dopant checklists without inventing a joint science.
Collapse: if Mg–H ≡ θ-role; if 2×10¹⁹ ≡ arity integer; if self-compensation ≡ adjunct licensing; if Navajo morphological typing is treated as evidence; if “electronic saturation” is treated as identical to precipitation or SSS saturation; if a joint field is announced.
Explicit non-equations:
- N_D / N_A / [Mg] / at.% ≠ θ-role fillers.
- Mott MIT ≠ over-arity as science identity.
- H activation / V_O anneal ≠ synchronic/diachronic linguistics (mnemonic only if used — collapse if elevated).
- Bridge A lock unrevised ( SPECULATIVE pedagogy / probe only). No joint field.
9. What to do next (reader checklist)
- Open the three host checklists; do not invent numbers past them.
- Name shallow vs deep vs mid-gap / V_O before comparing windows.
- Name which clock (implant/anneal; H activation; p(O₂)/V_O).
- List the plural “over” modes for that host — do not stop at “too much dopant.”
- Audit vocabulary (Path D): dopant here; never for percent-level metal SSS.
- Contrast foils briefly: intercalation · NMC “papers say doping” · Al–Mg SSS.
- Only then, if useful for memory, hang the SPECULATIVE Bridge A count peg.
- Cross-check Family 4 / HCPHexagonal close-packed — fewer primary slip systems; Mg, Ti, Zn often anisotropic. / SFE depth notes if the abstract said “doping” but the meter was strength or twinning.
10. Sources / corpus pointers
| Source | Role |
|---|---|
| Si:P/As/B | REAL: shallow P/As/B; Mott N_crit; implant/CZ; extrinsic windows |
| GaN:Mg | REAL: Mg_Ga depth; H activation; self-comp / PIDs; SIMS ≠ Hall p |
| Oxide electronic | REAL: TiO₂ mid-gap / V_O; AZO Al%; recombination / μ / secondary phases |
| Mechanism taxonomy §1 | Family 1 map |
| Physics of doping | Domain-pure doping ladder; oxide §3 caveats |
| CNT / graphene | Family 2 intercalation foil |
| NMC:Mg cathode | NMC “papers say doping” foil |
| Al–Mg | Al–Mg SSS foil |
| Doping vs alloying vocabulary | Path D vocabulary lock |
| Same-element Mg | GaN:Mg vs NMC:Mg vs Al–Mg |
| Precipitation family | Family 4 contrast (precip ≠ carrier) |
| Conclusions | Sixteen-host table; Bridge A lock |
| Reading order §2 | Electronic doping spine |
v0 — 2026-09-21 PT. Real electronic-doping contrast across Family 1 (Si · GaN:Mg · oxide); Bridge A flagged SPECULATIVE. Shallow ≠ deep ≠ mid-gap/V_O. Electronic ≠ intercalation ≠ NMC stabilizer ≠ Al–Mg SSS. No joint field.
Corpus note: depth-electronic-doping-family-contrast-v0.md. Cross-domain mappings are speculative.