driftlet 1.0.1 → 1.1.1

package/README.md

@@ -10,7 +10,7 @@
 
 **driftlet simulates charged species moving through one-dimensional devices**: semiconductor
 junctions and solar cells, electrochemical cells and batteries, double layers, membranes and
-nerves. It's a small JavaScript library with no dependencies, fast enough to solve live in a
+nerves. It's a JavaScript library with no dependencies, fast enough to solve live in a
 web page as you drag a slider.
 
 [Live demos](https://markblundeberg.github.io/driftlet/demos/) · [npm](https://www.npmjs.com/package/driftlet) ·
@@ -56,7 +56,7 @@ import { bandDiagram } from 'driftlet/plot';    // a level diagram as an SVG str
 ```
 
 Or straight from a CDN in a page, pinned to a version:
-`https://cdn.jsdelivr.net/npm/driftlet@1.0.1/src/index.js` (and `src/kit.js`, `src/plot.js`).
+`https://cdn.jsdelivr.net/npm/driftlet@1.1.1/src/index.js` (and `src/kit.js`, `src/plot.js`).
 TypeScript declarations ship in the package. Writing a demo with an LLM agent? Point it at
 [`llms.txt`](llms.txt).
 

package/docs/data.md

@@ -21,6 +21,12 @@ double layers. The data are in `IONS`, and $`\mu^\circ`$ of liquid water is `H2O
 1.011e-14 at 25 °C (pK_w 13.995): neutral water holds 1.0055e-4 mol/m³ of each, which a bath
 with that reaction should hold too.
 
+The data are 25 °C's, and the material says so (`T: 298.15`): a device at another temperature
+warns that they don't hold there. Its μ° (so equilibrium constants and electrode potentials),
+permittivity and diffusivities would all move: $`K_w`$ by ~6× between 6 and 25 °C. Scaling them
+(μ° by the standard entropies, ε, D by water's viscosity) is planned, not done. Where only
+transport between solutions of the same ions matters (a membrane's channels), μ° cancels.
+
 | Ion | z | D (10⁻⁹ m²/s) | $`\mu^\circ`$ (kJ/mol) |
 |---|---|---|---|
 | H+ | +1 | 9.311 | 0 |
@@ -53,16 +59,37 @@ with that reaction should hold too.
   Fe³⁺ and Fe²⁺ gives $`E^\circ(\mathrm{Fe}^{3+}/\mathrm{Fe}^{2+})`$ = 0.769 V, against the
   tabulated 0.771 V, which is what a Fe³⁺/Fe²⁺ redox demo needs.
 
-## Semiconductors, 300 K
+## Semiconductors, 300 K and other temperatures
 
-`semiconductor(name, { material })` returns electrons `e-` and holes `h+` and the material:
-$`c_{\mathrm{ref}} = N_c`$ and $`N_v`$, D from the mobilities by the Einstein relation at 300 K,
+`semiconductor(name, { material, T })` returns electrons `e-` and holes `h+` and the material:
+$`c_{\mathrm{ref}} = N_c`$ and $`N_v`$, D from the mobilities by the Einstein relation at T,
 $`\mu^\circ_{\mathrm{e}^-} = 0`$ (so $`\phi`$ sits at the conduction band) and
-$`\mu^\circ_{\mathrm{h}^+} = E_g \cdot F`$. The data are for 300 K, so give the device `T: 300`.
-Statistics are ideal (Boltzmann); add Fermi–Dirac statistics to the material for degenerate
-doping. The data are in `SEMICONDUCTORS`, in the source's units (cm⁻³, cm²/(V·s), eV).
+$`\mu^\circ_{\mathrm{h}^+} = E_g \cdot F`$. The table is for 300 K, the default; `T` (150 to 500 K)
+moves it as [below](#at-other-temperatures). Statistics are ideal (Boltzmann); add Fermi–Dirac
+statistics to the material for degenerate doping. The data are in `SEMICONDUCTORS`, in the
+source's units (cm⁻³, cm²/(V·s), eV). In the material they're in SI: $`N_c`$ and $`N_v`$ are the
+`cRef` of e⁻ and h⁺ in mol/m³, at `materials.Si.species['e-'].cRef` and `['h+'].cRef` (the
+`species` list carries only names and charges).
+
+**Give the device the same `T`.** The material's gap, $`N_c`$, $`N_v`$ and D are for the
+temperature it was made at, and it carries that temperature as `materials.Si.T`. A device at
+another temperature warns, in `sol.warnings` (so `check()`) and `describe()`, naming both
+temperatures. The device's own default is 298.15 K (25 °C, the ions' temperature), not the
+table's 300 K, so a semiconductor device says `T: 300`: at 298.15 K the 300 K data's
+intrinsic density is 13% low. A material written by hand has no `T` and is never warned about.
+
+To change the temperature of a live device, give it the material for the new T in the same
+`set()`. The state carries over as a warm start:
+
+```js nocheck
+dev.set({ T: 350, materials: semiconductor('Si', { T: 350 }).materials });
+```
+
+A `set({ T })` alone changes kT but not the material's data, and warns. (A material with a
+`trap()` in it: remake it with `trap(semiconductor('Si', { T }), …)` the same way. The trap
+takes its material's T.)
 
-| | $`\varepsilon_r`$ | $`E_g`$ (eV) | $`\chi`$ (V) | $`N_c`$ (cm⁻³) | $`N_v`$ (cm⁻³) | $`\mu_n`$ (cm²/(V·s)) | $`\mu_p`$ (cm²/(V·s)) |
+| | $`\varepsilon_r`$ | $`E_g`$ (eV) | $`\chi`$ (eV) | $`N_c`$ (cm⁻³) | $`N_v`$ (cm⁻³) | $`\mu_n`$ (cm²/(V·s)) | $`\mu_p`$ (cm²/(V·s)) |
 |---|---|---|---|---|---|---|---|
 | Si | 11.9 | 1.12 | 4.05 | 2.8 × 10¹⁹ | 1.04 × 10¹⁹ | 1500 | 450 |
 | Ge | 16.0 | 0.66 | 4.0 | 1.04 × 10¹⁹ | 6.0 × 10¹⁸ | 3900 | 1900 |
@@ -70,14 +97,108 @@ doping. The data are in `SEMICONDUCTORS`, in the source's units (cm⁻³, cm²/(
 
 - Source: Sze, *Physics of Semiconductor Devices*, 2nd ed. (1981), appendix "Properties of Ge,
   Si and GaAs at 300 K". The mobilities are for lightly doped material.
-- The electron affinity $`\chi`$ is for vacuum-level estimates with `vacuumDipole` (anchor `'e-'`,
-  see the [alignment guide](alignment.md)). It isn't used otherwise, since an alignment is never
-  implied.
+- The electron affinity $`\chi`$ (an energy, in eV) is for vacuum-level estimates with
+  `vacuumDipole` (anchor `'e-'`, `offset: chi`, the same number in volts; see the
+  [alignment guide](alignment.md)). It isn't used otherwise, since an alignment is never implied.
 - **Silicon's intrinsic concentration**: these $`N_c`$, $`N_v`$ and $`E_g`$ give $`n_i`$ = 6.7 × 10⁹ cm⁻³.
   Sze quotes 1.45 × 10¹⁰ (an older measurement), and the accepted value is 9.65 × 10⁹. The
   library keeps the band data self-consistent rather than matching a quoted $`n_i`$. For Ge and
   GaAs, the data reproduce the quoted $`n_i`$ (2.4 × 10¹³ and 1.79 × 10⁶) to within 10–15%.
 
+### At other temperatures
+
+With `T`, each 300 K number moves by a standard temperature law:
+
+- **Band gap**, by Varshni's form $`E_g(T) = E_{g0} - \alpha T^2/(T + \beta)`$ (Varshni, Physica 34,
+  149, 1967), with Thurmond's parameters as tabulated by Sze (Thurmond, J. Electrochem. Soc. 122,
+  1133, 1975). Only the shift from 300 K is used, on the table's gap,
+  $`E_g(T) = E_g^{\text{table}} + [E_g^{\text{V}}(T) - E_g^{\text{V}}(300\ \mathrm{K})]`$, so that 300 K is the
+  table exactly. Varshni's own value at 300 K differs from the table by a few meV.
+- **$`N_c`$ and $`N_v`$** $`\propto T^{3/2}`$, from parabolic bands with fixed effective masses.
+- **Mobilities**, by the power laws for scattering by lattice vibrations,
+  $`\mu \propto T^{-\gamma}`$ (Sze, ch. 1). They are approximate and hold for lightly doped material.
+  Where scattering by ionized impurities matters (heavy doping, or the cold end of the range), the real
+  mobility is lower and rises less on cooling. D is then $`\mu kT/q`$ at T, so it falls more slowly
+  than μ.
+- **Electron affinity** $`\chi`$ and **$`\varepsilon_r`$** stay at their 300 K values.
+
+| | $`E_{g0}`$ (eV) | $`\alpha`$ (10⁻⁴ eV/K) | $`\beta`$ (K) | Varshni at 300 K (eV) | $`\gamma_n`$ | $`\gamma_p`$ |
+|---|---|---|---|---|---|---|
+| Si | 1.170 | 4.73 | 636 | 1.1245 | 2.42 | 2.20 |
+| Ge | 0.7437 | 4.774 | 235 | 0.6634 | 1.66 | 2.33 |
+| GaAs | 1.519 | 5.405 | 204 | 1.4225 | 1.0 | 2.1 |
+
+Outside 150–500 K the call is refused: Varshni's fits and the lattice-limited mobilities are
+not meant for the cold, where impurities and freeze-out take over. With these laws, silicon's
+$`n_i`$ grows 8.3% per kelvin near room temperature. In the tests, a silicon cell's
+open-circuit voltage falls 2.14 mV/K at 300 K. That lies within Green's
+$`\mathrm{d}V_{oc}/\mathrm{d}T = -(E_{g0}/q - V_{oc} + \gamma kT/q)/T`$ for the 1.5 ≤ γ ≤ 3 that
+its transport allows, and the γ = 1.67 it implies matches its two parts of $`J_0`$.
+
+**A diode's saturation current** follows from these laws. With $`n_i^2 \propto T^3 e^{-E_g/kT}`$
+and $`D = \mu kT/q \propto T^{1-\gamma}`$ ($`\gamma`$ of the minority carrier that dominates),
+a short base, $`J_0 \propto n_i^2 D/W`$, gives
+
+```math
+J_0 \propto T^{3 + 1 - \gamma}\, e^{-E_g(T)/kT}, \qquad E_a = -k\,\frac{\mathrm{d}\ln J_0}{\mathrm{d}(1/T)} = E_g - T\frac{\mathrm{d}E_g}{\mathrm{d}T} + (4 - \gamma)\,kT .
+```
+
+So an Arrhenius plot of $`J_0`$ doesn't read the gap. $`E_g - T\,\mathrm{d}E_g/\mathrm{d}T`$ is
+the gap extrapolated linearly to 0 K, 1.196 eV for silicon at 300 K, and the prefactor adds
+$`(4-\gamma)kT`$, 0.041 eV for its electrons, about 1.24 eV in all. A long base,
+$`J_0 \propto n_i^2\sqrt{D/\tau}`$, has $`(1-\gamma)/2`$ in place of $`1-\gamma`$, with τ as you
+give it (the library doesn't move it with T).
+
+**Hot, lightly doped junctions disappear.** $`n_i`$ grows fast: silicon's is 3 × 10¹³ cm⁻³ at
+450 K and 1.6 × 10¹⁴ at 500 K, germanium's 6 × 10¹⁵ at 450 K and 2 × 10¹⁶ at 500 K. Where $`n_i`$
+approaches the doping, both sides go intrinsic and the junction (its built-in voltage, its
+rectification) fades away. Compare $`n_i(T)`$ with the doping before reading a trend in V(T)
+across that range.
+
+### Alloys and other materials from these
+
+A material the library lacks can start from a near relative's entry, with its own numbers put
+in. For example, Al₀.₃Ga₀.₇As from GaAs: give the holes' `mu0` the alloy's gap ($`E_g F`$), and set
+whatever else differs, such as the mobilities and $`N_c`$, $`N_v`$. A field you don't set keeps
+GaAs's value, its electron mobility of 8500 cm²/(V·s) included. Then give each face between the
+two its band offset as a `step` on e⁻, $`\Delta E_c`$ right minus left (the valence-band offset
+follows from the two gaps). `semiconductor(name, { T })` moves only the named material's own data,
+by its laws: numbers you put in stay as given, so at another temperature move them yourself. The
+material still carries the `T` it was made at, for the warning.
+
+```js
+import { Device, units, GAS_CONSTANT, FARADAY } from 'driftlet';
+import { build, layer, ohmic, semiconductor } from 'driftlet/kit';
+
+// The usual estimates for x = 0.3 (E_g = 1.424 + 1.247x eV, ΔE_c ≈ 0.62 ΔE_g,
+// m_e = 0.067 + 0.083x); check them against your source.
+const T = 300, VT = (GAS_CONSTANT * T) / FARADAY, x = 0.3;
+const gap = 1.424 + 1.247 * x, dEc = 0.62 * (gap - 1.424);
+const alloy = semiconductor('GaAs', { material: 'AlGaAs', T }), m = alloy.materials.AlGaAs;
+m.species['h+'] = { ...m.species['h+'], mu0: units.eV(gap) };
+m.species['e-'] = {
+  ...m.species['e-'],
+  D: 1000e-4 * VT, // μ_n = 1000 cm²/(V·s), a rough figure
+  cRef: m.species['e-'].cRef * ((0.067 + 0.083 * x) / 0.067) ** 1.5, // N_c ∝ m_e^1.5
+};
+// A heterojunction diode, n-AlGaAs | p-GaAs, forward biased.
+const dev = new Device(build({
+  T,
+  library: [semiconductor('GaAs', { T }), alloy],
+  stack: [
+    ohmic(0), layer('AlGaAs', 1e-6, { donors: units.perCm3(1e17) }),
+    { step: { species: 'e-', value: units.eV(-dEc) } }, // GaAs's conduction band is lower
+    layer('GaAs', 1e-6, { acceptors: units.perCm3(1e17) }), ohmic(1.0),
+  ],
+  bulkReactions: [{ equation: 'e- + h+ = 0', kf: { GaAs: 1e8, AlGaAs: 1e8 } }],
+}));
+const sol = dev.solve();
+// The wider gap keeps holes out of the n side: nearly all the current is electrons injected
+// into the GaAs.
+const { 'e-': e, 'h+': h } = sol.contacts.left.flux;
+console.log(sol.warnings.length, (e / -h).toExponential(1)); // 0, 2.5e+7
+```
+
 ## Metals, 20 °C
 
 `metal(name, { material })` returns a conductor material carrying `e-`, with its conductivity

package/docs/device.md

@@ -15,7 +15,7 @@ too (`FARADAY`, `GAS_CONSTANT`, `EPS0`, …).
 
 | Key | Meaning |
 |---|---|
-| `T` | temperature, K (default 298.15) |
+| `T` | temperature, K (default 298.15, 25 °C; the library's semiconductors are for 300 K, so give them `T: 300`) |
 | `species` | list of mobile species |
 | `materials` | named materials: bulk properties |
 | `regions` | the device, left to right: each a material plus length, fixed charge, composition |
@@ -55,6 +55,11 @@ Per species present: `D` (m²/s, the coefficient in the flux $`N = -(D c/RT)\nab
 (standard chemical potential, J/mol) and `cRef` (the concentration that `mu0` refers to; falls back
 to the species' `cRef`, and must exist one way or the other). Species not listed are absent.
 
+A material may say what temperature its numbers are for, `T` (K). `semiconductor()` sets it.
+The device then warns, in its solutions' `warnings`, if that isn't its own `T` (to 0.01 K), since
+the material's gap, densities of states and D are then another temperature's. A material without
+`T` is taken as written.
+
 In a material, only *neutral combinations* of standard potentials mean anything physically.
 Shifting every `mu0` by $`z_i F s`$ just moves the material's $`\phi`$, its bookkeeping anchor. See
 [conventions](conventions.md).
@@ -245,7 +250,9 @@ r = \frac{n\,p\,(1 - e^{-A/RT})}{(n + n_1)/v_p + (p + p_1)/v_n}, \qquad p_1 = \f
 with n and p each at its own side's edge node, the capture velocities $`v_n`$ and $`v_p`$ in m/s,
 and $`n_1`$ (mol/m³) the negative species' concentration with its level at the trap (default:
 $`n_1 = p_1 = n_i`$, a midgap trap). Taking $`p_1`$ from the state keeps equilibrium exact whatever
-lies between the two sides (a band offset, a $`\phi`$ jump). Where one carrier is plentiful the
+lies between the two sides (a band offset, a $`\phi`$ jump). With Fermi–Dirac carriers, each
+one's emission scales by its own $`\gamma = \mathcal{F}_{1/2}(\zeta)/e^\zeta`$, by detailed balance:
+$`n_1`$ becomes $`\gamma_n n_1`$, and $`p_1`$ keeps the product. Where one carrier is plentiful the
 rate saturates at the other's capture, $`r \to v_p\,p`$: holes reaching an electron-rich layer
 recombine at $`v_p`$ however many electrons wait there, which plain mass action can't do.
 
@@ -418,7 +425,13 @@ those materials.
 none (`'e- + h+ = 0'`):
 $`r = n p (1 - e^{-A/RT}) / (\tau_p (n + n_1) + \tau_n (p + p_1))`$, with lifetimes in s and $`n_1`$ as
 for [interface SRH](#interfaces) (default midgap). It's what explicit trap species reduce to when
-the traps are few and fast, without them.
+the traps are few and fast, without them. Where their charge matters, write them out: two
+immobile species (D = 0) for the trap's charge states, filled and emptied by
+`'e- + X0 = X-'` and `'X- + h+ = X0'` (acceptor-like) at rates $`k_n = 1/(\tau_n N_t)`$ and
+$`k_p = 1/(\tau_p N_t)`$, the filled state's $`\mu^\circ`$ an electron's $`\mu`$ at $`n_1`$ above the
+empty one's. In steady state they recombine at the rate above exactly, each node keeps its
+own trap total, and the trapped charge is in Poisson's equation. The kit's
+[`trap()`](kit.md#a-device-as-a-stack) writes them from the same `{ tauN, tauP, n1 }` and a density.
 
 ```js nocheck
 bulkReactions: [{ equation: 'e- + h+ = 0', srh: { Si: { tauN: 1e-6, tauP: 1e-6 } } }]
@@ -1000,6 +1013,9 @@ const now = dev.solution();                   // snapshot of the current state
   mid-run, `set({ contacts: { right: { bath: { c } } } })`, is a step in the sample.) Either way the time stepping restarts its
   order, as after any discontinuity, but a carried state goes on at the step size an adaptive
   transient had reached, so a page may change a rate constant or a diffusivity every frame.
+  To change the temperature, give the materials for it in the same patch,
+  `dev.set({ T: 350, materials: semiconductor('Si', { T: 350 }).materials })`: a `T` alone
+  changes kT but not the materials' data (a material that carries its `T` then warns).
 
 For example, a silver nitrate cell between silver electrodes: its impedance spectrum, then the
 current transient after a voltage step.
@@ -1047,7 +1063,7 @@ console.log(`${run.steps} steps; I(0.01 s) ≈ ${run.trace.current[run.trace.t.f
 | `bulkReactions[k]` | `{ rate, forward, nu, regions, total }`: each bulk reaction's net forward rate at every node (mol/(m³·s), `NaN` where it doesn't run), its one-way forward rate (the backward is `forward − rate`), what one forward reaction makes of each species (`nu`, negative for what it consumes), and integrated over each region and the device (mol/(m²·s)), as the balances count it: a charge-balance check is J = F(generation − recombination), which the kit's `check()` does for every species |
 | `charge` | total charge in the device, C/m² (C through a cross-section) |
 | `conservation` | per species stretch: amount, reference, intake through contacts, and drift, relative: (amount − reference − intake) over the largest of them |
-| `warnings` | e.g. unresolved double layers, conventions a statistics model relies on, and for a failed solve, where the system is nearly singular |
+| `warnings` | e.g. unresolved double layers, conventions a statistics model relies on, a material's data for another temperature, and for a failed solve, why (a driven current it can't pass, where the system is nearly singular). A failed steady solve's state isn't a solution, so the warnings about a solution's accuracy (the grid, double layers, space charge reaching a contact) are left out of it |
 | `converged`, `iterations`, `steps`, `substeps`, `history`, `time`, `T` | solver bookkeeping, and the temperature |
 | `steady` | whether this is a converged steady state (from `solve()`), not a transient's: what `check()` in the kit decides its checks by |
 | `done`, `rejected`, `trace` | from `advance()`: whether `tEnd` was reached; rejected steps; `{ t, current, voltage }` per accepted step |

package/docs/kit.md

@@ -68,6 +68,15 @@ console.log(`I = ${sol.current.toFixed(1)} A/m²`);
 - **`library`** takes pieces with `species` and `materials`, merged with any given directly
   (`combine()` does the merging on its own). A species or material given twice must be given
   identically.
+- **`trap(lib, { density, tauN, tauP, n1, charge, name, material, T })`** adds a trap level that
+  holds charge to a semiconductor's library piece: its two states as immobile species
+  (`X0`/`X-` for `charge: 'acceptor'`, the default; `X+`/`X0` for `'donor'`), filled and emptied
+  by capturing electrons and holes. It returns `{ library, bulkReactions, c0 }`: the library to
+  give `build()` in place of `lib`, the two capture reactions, and the trap's total for each
+  layer that has it (`layer('Si', L, { donors, c0: t.c0 })`). In steady state it recombines at the
+  `srh` law's rate with the same `tauN`, `tauP` (s, at that density) and `n1` (mol/m³, default
+  n_i); unlike `srh`, its charge is in Poisson's equation
+  ([device](device.md#bulk-reactions)).
 - Nested lists in the stack are flattened, so a helper can return several items.
 - Reactions pass through as written (see below), and so do `T`, `grid`, `ports` and `geometry`.
 
@@ -395,7 +404,11 @@ warnings:
   `warnings` check this again against the solved concentrations);
 - an outside (a bath, a contact's levels) out of equilibrium with a bulk reaction that runs at the
   contact, which would run there for ever: H⁺ and OH⁻ in a bath a little off $`K_w`$ pass a current
-  at zero bias.
+  at zero bias;
+- the definition's own warnings, which every solution repeats: a material whose data are for
+  another temperature than the device's (`semiconductor('Si')` is 300 K's; the device's default
+  is 298.15 K), a starting profile the grid doesn't resolve, conventions a statistics model
+  relies on.
 
 They're heuristics, prompts to check, never errors.
 
@@ -424,14 +437,17 @@ page can show that its numbers hold up. Pass the solution `solve()` or `advance(
 
 - **converged**: the solve converged.
 - **warnings**: the solution's warnings (unresolved double layers, steep profiles, a weakly held
-  part) and the definition's likely unit slips. They're heuristics, so they're marked `?`, a
+  part, a material's data for another temperature) and the definition's likely unit slips. They're heuristics, so they're marked `?`, a
   prompt to look, and don't fail the check.
 - **balance**, in a steady state: each species' ledger, what comes in through each terminal and
   what each reaction (bulk, or at a face) makes or uses, in mol/(m²·s) and for a charged species
   as a current. The terms must sum to zero. A species that a face lets through only by a law (a
   membrane's permeability, a conductance) or by a reaction has a ledger on each side, with what
   crosses as a term: in a cell, Na⁺ leaks into the cytoplasm as fast as the pump puts it out. Under light it's J = F(G − R), and it says where
-  every carrier went: collected, or recombined in the bulk or at which face. A species at rest
+  every carrier went: collected, or recombined in the bulk or at which face. Each reaction is a
+  term named by its equation; two with the same equation (radiative and SRH recombination, both
+  `e- + h+ = 0`) are told apart by their place, `e- + h+ = 0 (bulkReactions[1])`, and
+  `sol.bulkReactions[k]` has each one's rates. A species at rest
   (in equilibrium, flows only of round-off against what it could carry) isn't listed.
 - **conservation**, in a transient: every closed, unreacting stretch keeps what it held plus
   what came in.

package/docs/numerics.md

@@ -450,8 +450,8 @@ The assembled matrix loses what a flux's two entries share. A Scharfetter–Gumm
 $`\eta`$ at its two ends through $`G`$ and $`-G(1 + E)`$, with $`E = \mathrm{expm1}(\Delta\eta)`$;
 for a nearly uniform $`v`$ they cancel, and $`J \cdot v`$ keeps only round-off of $`G|v|`$, which in
 an inversion layer's 0.1 nm cells ($`G`$ ~ 1e11) dwarfs the flux. Where a solve needs
-$`J \cdot v`$ itself (GMRES, in the impedance and in Newton's refined solves), the dilute kernels'
-terms are kept aside instead, each evaluated on differences first:
+$`J \cdot v`$ itself (GMRES, in the impedance and in Newton's refined solves), the kernels' terms
+are kept aside instead, each evaluated on differences first:
 
 ```math
 \delta N = a\,(\delta\eta_L - z\,\delta\hat\phi_L) + b\,(\delta\eta_R - \delta\eta_L) + d\,(\delta\hat\phi_R - \delta\hat\phi_L),
@@ -459,8 +459,13 @@ terms are kept aside instead, each evaluated on differences first:
 
 with $`a = -GE`$ (through $`c_L`$), $`b = -G(1 + E)`$ and $`d`$ the drift's. A node's storage and space
 charge depend on $`\delta\eta_i - z_i\,\delta\hat\phi`$ alone, the displacement on
-$`\delta\hat\phi_R - \delta\hat\phi_L`$. Everything else (faces, contacts, ports, reactions,
-concentrated and metal regions) is assembled into a matrix of its own. Rows that a contact or a
+$`\delta\hat\phi_R - \delta\hat\phi_L`$. Non-ideal statistics couple the species ($`c_i`$
+depends on every $`\zeta_j`$), so there a node's storage of $`i`$ has a term in each
+$`\delta\eta_j - z_j\,\delta\hat\phi`$, through $`K_{ij}`$, and a flux has side terms in those at
+either end (through $`c_L`$ and the excess potentials in its drift); each is in the same
+differences. (Until 1.1.1 they were assembled directly, and a Fermi–Dirac MOS capacitor in
+inversion, barely degenerate, came out with a negative resistance.) Everything else (faces,
+contacts, ports, reactions, metal regions) is assembled into a matrix of its own. Rows that a contact or a
 steady solve replaces drop the kept terms; the transformed rows of strictly neutral nodes combine
 them as they combine the matrix's. A region's uniform shift ($`z_i s`$ on each $`\eta_i`$, $`s`$ on
 $`\hat\phi`$) then gives exactly zero inside it, as it must. It replaced a central difference of
@@ -649,7 +654,7 @@ majority conductances times a $`\delta\eta`$ cancelling its neighbours' to 1e-17
 round-off reads as a constant conductance of order 1 S/m², a silicon p⁺n⁺ junction's capacitance
 9× off at 100 Hz. Between the two contacts of a device without ports, the total current
 (conduction plus displacement, $`i\omega\,\delta D`$) is the same through every cut, so it is also
-read across each segment of an ideal region, linearised as the Scharfetter–Gummel flux is, and
+read across each segment of a region (with any statistics), linearised as its flux is, and
 the reading with the least estimated error is taken: in that junction, the depletion region's;
 in a MOS capacitor, the oxide's displacement. A reading's error is its terms' round-off, and the
 solves' error in $`\delta x`$ (their residual, relative to its largest entry) through its

package/docs/reliability.md

@@ -58,6 +58,7 @@ recipes.
 | SRH recombination | bulk: G = R_SRH(n, p) with n = N_D + p, midgap and shallow traps, low to high injection; at a face: zero in the dark, the law between the two edge nodes, every pair collected or recombined | 1e-6; 1e-9 |
 | Perovskite J–V hysteresis, against IonMonger | an independent code (finite elements, Octave): mobile iodide vacancies, interface SRH, scans from 1 mV/s to 1 kV/s — hysteresis index, P_max, V_oc, the whole loop (as current where it's gentle, as a voltage offset near V_oc, where it falls at up to 1100 mA/cm² per volt) | 1.5e-3; 0.05 mW/cm²; 2 mV; 0.03 mA/cm² and 1 mV (0.010 and 0.31 mV found) |
 | Illuminated long pn diode | `J_sc = qG(L_n + L_p + W)` from a cold start; superposition at low injection | 5e-3; 1e-2 |
+| A silicon cell's V_oc against temperature (`semiconductor('Si', { T })`) | Green's dV_oc/dT = −(E_g0/q − V_oc + γkT/q)/T for 1.5 ≤ γ ≤ 3 (−2.12 to −2.25 mV/K); the γ it implies against its two parts of J₀, 1.66; n_i's growth, 8.3%/K, against 3/2T + E_g0/2kT² | within the band (−2.14 found); ±0.1 (1.67 found); 1e-4 |
 | Schottky barrier (metal region \| n-Si) | surface density from the alignment; depletion charge | 5e-3; 2% |
 | Surface velocity at a contact | minority electrons through a base to a contact of velocity S, n₀(e^(V/V_T) − 1)/(W/D + 1/S), S from 10 to 1e5 m/s; a Schottky diode's thermionic emission, short of A*T²e^(−φ_B/V_T)(e^(V/V_T) − 1) by its depletion region in series, the same swept as built cold | 1e-5; 3–7%; 1e-6 |
 | Charged pore or nanochannel (resolved Donnan layers at each mouth) | conductance against salt, down to its plateau, as Teorell–Meyer–Sievers in a long pore (short ones short by end layers ∝ 1/L); the zero-current voltage across a salt gradient; a symmetric pore's I–V odd, rectification with charge on one side or bipolar | 1%; 5e-3 to 1e-4; 1e-9 |
@@ -172,9 +173,6 @@ so a change that makes one worse shows.
 - **A surface covered to a bare fraction below ~1e-10** barely affects anything any more, and
   the solver struggles with it (the hard corpus's passivating drop without a double layer stops
   partway through its hour); the solution warns, and suggests a less stable film.
-- **A double layer across more than ~350 thermal voltages** (between 9 and 9.25 V at 25 °C, in a
-  crowded lattice-gas electrolyte) doesn't solve, from a cold start or warm, and its warning
-  doesn't say why. Ionic-liquid windows (a few volts) are well inside it.
 - **Currents below ~0.1 nA/m²** (a GaAs junction's leakage without recombination) aren't resolved
   between solves; such a current is zero for any practical purpose.
 - **A failed solve says so:** `converged: false`, with a warning naming where the system lost

package/docs/templates.md

@@ -214,7 +214,7 @@ for (let k = 1; k <= 300; k++) {
   const sol = axon.advance(k * 1e-4); // 0.1 ms at a time
   peak = Math.max(peak, sol.interfaces[0].V); // V_m, and sol.interfaces[0].gates = { m, h, n }
 }
-console.log(`peak ${(peak * 1e3).toFixed(1)} mV`); // +12.6: GHK's Na⁺ current is limited by the 50 mM inside (linear: true gives +37.9, HH's own)
+console.log(`peak ${(peak * 1e3).toFixed(1)} mV`); // +12.6: GHK's currents rectify; matched at rest, at positive V Na⁺'s conductance is lower and K⁺'s higher (linear: true gives +37.9, HH's own)
 ```
 
 ## an action potential travelling along an axon

package/llms.txt

@@ -30,9 +30,9 @@ materials, insertion hosts), data.md and reliability.md (what's validated, and t
 From a CDN, pin the version you tested against:
 
 ```js nocheck
-import { Device, units } from 'https://cdn.jsdelivr.net/npm/driftlet@1.0.1/src/index.js';
-import { build, layer, ohmic, semiconductor, live } from 'https://cdn.jsdelivr.net/npm/driftlet@1.0.1/src/kit.js';
-import { bandDiagram } from 'https://cdn.jsdelivr.net/npm/driftlet@1.0.1/src/plot.js';
+import { Device, units } from 'https://cdn.jsdelivr.net/npm/driftlet@1.1.1/src/index.js';
+import { build, layer, ohmic, semiconductor, live } from 'https://cdn.jsdelivr.net/npm/driftlet@1.1.1/src/kit.js';
+import { bandDiagram } from 'https://cdn.jsdelivr.net/npm/driftlet@1.1.1/src/plot.js';
 ```
 
 From npm: `driftlet`, `driftlet/kit`, `driftlet/plot`. Every export is listed at the end.
@@ -73,12 +73,16 @@ From npm: `driftlet`, `driftlet/kit`, `driftlet/plot`. Every export is listed at
   are regions (give one a `name` to refer to it), and any other object between two layers is
   the face between them. `ohmic()`, `bath()` and the like write contacts; a plain object works.
 - **The library** is pieces with `species` and `materials`: `semiconductor('Si')` (Si, Ge, GaAs, at
-  300 K), `aqueous(names)` (H⁺ Li⁺ Na⁺ K⁺ Ag⁺ Cu²⁺ Zn²⁺ Fe²⁺ Fe³⁺ OH⁻ Cl⁻ NO₃⁻ SO₄²⁻ in water, at 25
+  300 K; `semiconductor('Si', { T: 350 })` moves them, 150–500 K, for a device at that `T`),
+  `aqueous(names)` (H⁺ Li⁺ Na⁺ K⁺ Ag⁺ Cu²⁺ Zn²⁺ Fe²⁺ Fe³⁺ OH⁻ Cl⁻ NO₃⁻ SO₄²⁻ in water, at 25
   °C; `{ epsr: 0 }` makes it strictly neutral, with no double layers), `metal('Ag')` (Ag Cu Au Al Zn
   Li Pt), all from `driftlet/kit`. A semiconductor's textbook numbers are in `SEMICONDUCTORS.Si`
   (`Eg` and `chi` in eV, `Nc` and `Nv` in cm⁻³, `mun` and `mup` in cm²/(V·s), `epsr`); in the
-  material, e⁻'s and h⁺'s `cRef` are N_c and N_v in mol/m³ and h⁺'s `mu0` is the gap (e⁻'s is 0),
-  so n_i = √(N_c N_v) e^(−E_g/2RT), 6.7e9 cm⁻³ for Si. A species it lacks is pushed into both lists
+  material, `materials.Si.species['e-'].cRef` and `['h+'].cRef` are N_c and N_v in mol/m³ and
+  h⁺'s `mu0` is the gap (e⁻'s is 0), so n_i = √(N_c N_v) e^(−E_g/2RT), 6.7e9 cm⁻³ for Si. The
+  material carries the T its data are for (`materials.Si.T`), and a device at another T warns
+  (its default is 298.15 K, so say `T: 300`). An alloy starts from a relative's entry with your
+  numbers put in (docs/data.md has AlGaAs from GaAs). A species it lacks is pushed into both lists
   (`z: 0` for a neutral one): `lib.species.push({ name: 'A-', z: -1, cRef: 1000 })`,
   `lib.materials.water.species['A-'] = { D, mu0 }`. Only μ° differences within a reaction matter:
   for a redox couple μ°_red − μ°_ox = −F·E° (against SHE, the library's μ°_H⁺ being 0). An
@@ -115,7 +119,11 @@ From npm: `driftlet`, `driftlet/kit`, `driftlet/plot`. Every export is listed at
   bimolecular, as is the kit's `recombination()`, it shortens at high injection. SRH is `srh: { Si:
   { tauN, tauP } }` in place of `kf` (s), or at a face `{ equation: 'e-(left) + h+(right) = 0',
   srh: { vn, vp } }` (m/s); with τ_N ≪ τ_p it gives n-type silicon's holes a lifetime τ_p at any
-  injection. Light is a reaction from a reservoir (`photon = e- + h+`, uniform), or absorbed as
+  injection. `srh` carries no charge; for traps whose charge matters (dense or deep ones, trapped
+  charge bending the bands), `trap(semiconductor('Si'), { density, tauN, tauP, n1, charge:
+  'acceptor' })` writes the trap's two states as immobile species and returns `{ library,
+  bulkReactions, c0 }` (c0 into each layer that has the trap): the same rate in steady state, its
+  charge in Poisson. Light is a reaction from a reservoir (`photon = e- + h+`, uniform), or absorbed as
   it goes in, `photogeneration({ material, flux, alpha, mu: units.eV(3), from, to })`, `from` and
   `to` being x in the device (a port's window is instead m from its region's left end). To switch
   it during a run, `set()` the device's whole `bulkReactions` list, recombination included: in
@@ -168,7 +176,7 @@ import { Device, units } from 'driftlet';
 import { build, layer, ohmic, semiconductor, describe } from 'driftlet/kit';
 
 const def = build({
-  T: 300, // the library's semiconductor data are for 300 K
+  T: 300, // the library's semiconductor data are for 300 K, unless semiconductor('Si', { T })
   library: [semiconductor('Si')],
   stack: [
     ohmic(0), // e⁻ and h⁺ in equilibrium with a metal at 0 V
@@ -176,7 +184,9 @@ const def = build({
     layer('Si', units.um(1), { name: 'p', acceptors: units.perCm3(1e16) }),
     ohmic(0.5), // the p side at +0.5 V: forward bias
   ],
-  bulkReactions: [{ equation: 'e- + h+ = 0', kf: { Si: 1e8 } }], // recombination: τ = 1/(kf·N) ≈ 0.6 µs at N = 1e16 cm⁻³
+  // recombination: τ = 1/(kf·N) ≈ 0.6 µs at N = 1e16 cm⁻³, so L = √(Dτ) ≈ 50 µm ≫ 1 µm: a short
+  // diode, whose Shockley J₀ takes each side's quasi-neutral width (W − x_d) in place of L
+  bulkReactions: [{ equation: 'e- + h+ = 0', kf: { Si: 1e8 } }],
   grid: { hmin: units.nm(1), hmax: units.nm(20) },
 });
 console.log(describe(def));
@@ -209,9 +219,9 @@ a CDN needs a long virtual-time budget, 30 s.)
 <span id="readout"></span>
 <div id="figure"></div>
 <script type="module">
-import { units } from 'https://cdn.jsdelivr.net/npm/driftlet@1.0.1/src/index.js';
-import { build, layer, ohmic, semiconductor, live } from 'https://cdn.jsdelivr.net/npm/driftlet@1.0.1/src/kit.js';
-import { bandDiagram } from 'https://cdn.jsdelivr.net/npm/driftlet@1.0.1/src/plot.js';
+import { units } from 'https://cdn.jsdelivr.net/npm/driftlet@1.1.1/src/index.js';
+import { build, layer, ohmic, semiconductor, live } from 'https://cdn.jsdelivr.net/npm/driftlet@1.1.1/src/kit.js';
+import { bandDiagram } from 'https://cdn.jsdelivr.net/npm/driftlet@1.1.1/src/plot.js';
 
 const def = build({
   T: 300,
@@ -641,8 +651,13 @@ those:
 - **Unit slips.** D in cm²/s, μ° in eV, µF/cm², mS/cm², T in °C. `describe(def)` flags them.
 - **A couple's μ° with the wrong sign** solves happily at −E° for +E°: μ°_red − μ°_ox = −F·E°.
   Check the rest potential against E°, as the stirred-solution template does.
-- **The library's data are for 25 °C** (its semiconductors for 300 K: give `T: 300`). At body
-  temperature the same μ° and D are a fair start; D rises about 2% per kelvin in water.
+- **The library's data are for 25 °C** (its semiconductors for 300 K: give `T: 300`, or the
+  same `T` to `semiconductor(name, { T })`, which moves the gap, N_c, N_v and the mobilities; a
+  mismatch warns). On a live device change both at once, `dev.set({ T, materials:
+  semiconductor('Si', { T }).materials })`. A diode's J₀ ∝ T^(4−γ) e^(−E_g/kT), so its Arrhenius
+  slope is E_g − T dE_g/dT + (4−γ)kT (≈ 1.24 eV for Si), not E_g. Hot and lightly doped, n_i
+  passes the doping (Ge 6e15 cm⁻³ at 450 K) and the junction fades: compare before fitting V(T).
+  At body temperature the same μ° and D are a fair start; D rises about 2% per kelvin in water.
 - **A steady solve of a closed cell where every ion can cross**: its steady state is complete
   mixing. To watch a junction or a salt bridge, `advance()` from the unmixed `c0`.
 - **A sharp junction between very different solutions** (3 M KCl against 1 µM) can't start:
@@ -663,7 +678,7 @@ those:
 - `driftlet/kit`:
   - building: `build`, `combine`, `layer`, `ohmic`, `bath`, `half`, `level`, `SHE`,
     `vacuumLevel`, `vacuumDipole`, `vacuumZeroCharge`, and the data library (`aqueous`,
-    `semiconductor`, `metal`, `IONS`, `SEMICONDUCTORS`, `METALS`, `H2O`, `WATER_EPSR`);
+    `semiconductor`, `metal`, `trap`, `IONS`, `SEMICONDUCTORS`, `METALS`, `H2O`, `WATER_EPSR`);
   - sources: `pulse`, `square`, `triangle`, `ramp`, `injector`, `recombination`,
     `photogeneration`;
   - channels: `hodgkinHuxley`, `ghkCurrent`;
@@ -674,4 +689,4 @@ those:
     `particles`;
   - curated devices: `perovskiteCell`, `PEROVSKITE_SCANS`, `hysteresis`.
 - `driftlet/plot` (a quick look): `bandDiagram`, `levelChart`, `labelParts`, `themeColor`,
-  `THEME`.
+  `THEME`. These draw profiles in x; an I–V curve or a transient you draw yourself.

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "driftlet",
-  "version": "1.0.1",
+  "version": "1.1.1",
   "description": "Dependency-free 1D drift–diffusion–reaction (Poisson–Nernst–Planck) solver for semiconductors, electrochemistry and nerves, fast enough for live web demos",
   "type": "module",
   "exports": {

package/src/checks.js

@@ -134,8 +134,12 @@ function balance(device, sol) {
     if (through('right', i)) add(i, last, 'right contact', -sol.contacts.right.flux[sp.name]);
     model.ports.forEach((port, k) => add(i, port.region, `port ${sol.ports[k].name}`, sol.ports[k].flux[sp.name]));
   });
+  // (two reactions with one equation, radiative and SRH recombination say, keep a line each,
+  // named by their place in the list)
+  const unique = (labels, where) => labels.map((w, k) => (labels.indexOf(w) === labels.lastIndexOf(w) ? w : `${w} (${where}[${k}])`));
+  const bulkWhat = unique(model.reactions.map((_, k) => label(def.bulkReactions?.[k])), 'bulkReactions');
   model.reactions.forEach((rx, k) => {
-    const what = label(def.bulkReactions?.[k]);
+    const what = bulkWhat[k];
     sol.bulkReactions[k].regions.forEach((total, r) => {
       const all = [...rx.reactants, ...rx.products].map((p) => p.i);
       for (const { i, nu } of rx.reactants) add(i, r, what, -nu * total, all);
@@ -153,8 +157,9 @@ function balance(device, sol) {
       add(i, f, `across face ${f} (${type})`, -N);
       add(i, f + 1, `across face ${f} (${type})`, N);
     });
+    const faceWhat = unique(itf.reactions.map((_, k) => label(def.interfaces?.[f]?.reactions?.[k])), 'reactions');
     itf.reactions.forEach((rx, k) => {
-      const rate = Af * sol.interfaces[f].rates[k], what = `${label(def.interfaces?.[f]?.reactions?.[k])} at face ${f}`;
+      const rate = Af * sol.interfaces[f].rates[k], what = `${faceWhat[k]} at face ${f}`;
       const all = rx.part.map((p) => p.i);
       for (const { i, nu, side } of rx.part) add(i, f + side, what, nu * rate, all);
     });