Struct FullPenn 

pub struct FullPenn { /* private fields */ }

The full Penn model built from one optical ELF. See the module docs.

Implementations

impl FullPenn

pub fn new(elf: OpticalElf) -> Self

The model for elf, with Fermi energy zero and the default tolerance.

pub fn with_fermi_energy_ev( self, fermi_ev: f64, ) -> Result<Self, ElectronDataError>

Set the Fermi energy (eV, finite and non-negative), as for the single-pole model: energies are kinetic energies above the Fermi level and the kinematics use T' = T + E_F.

pub fn with_relative_tolerance( self, rel_tol: f64, ) -> Result<Self, ElectronDataError>

Set the relative integration tolerance, in [1e-9, 1e-2].

pub fn optical_elf(&self) -> &OpticalElf

The optical ELF the model is built from.

pub fn fermi_energy_ev(&self) -> f64

The Fermi energy, eV.

pub fn relative_tolerance(&self) -> f64

The relative integration tolerance.

pub fn f_sum_ev2_at(&self, q_per_m: f64) -> Result<f64, ElectronDataError>

∫ dω_p g(ω_p) [∫ ω^j Im[-1/ε_L(q, ω)] dω over (0, ∞)] is the optical f-sum at j = 1: the ω integral of ω Im[-1/ε(q, ω)] at momentum q (m⁻¹), eV², evaluated Lindhard gas by Lindhard gas. The integral over the Lindhard continuum is cut off at ω_max = 40 (q k_F + q²/2 + ω_p) for each gas. The optical f-sum ∫ ω Im[-1/ε(ω)] dω of the same table (the value this must equal, super::sum_rules::SumRuleReport::f_sum_ev2) is independent of q.

pub fn loss_function(&self, q_per_m: f64, loss_ev: f64) -> f64

Im[-1/ε(q, ω)] of the full Penn algorithm (S2017 eq. (4)), at momentum transfer q (m⁻¹) and energy loss ω (eV). At q = 0 it is the optical ELF.

pub fn loss_function_evaluation( &self, q_per_m: f64, loss_ev: f64, ) -> LossFunctionEvaluation

Self::loss_function with the work of its ω_p integral (S2017 eq. (4)) and whether that integral met its tolerance; see LossFunctionEvaluation. The value is the same number Self::loss_function returns.

pub fn diimfp_per_m_ev( &self, energy_ev: f64, loss_ev: f64, ) -> Result<f64, ElectronDataError>

The DIIMFP p(T, ω) (S2017 eq. (2), nonrelativistic) at kinetic energy energy_ev above the Fermi level and energy loss loss_ev, in m⁻¹ eV⁻¹. Zero for ω <= 0 or ω > T.

pub fn diimfp_evaluation( &self, energy_ev: f64, loss_ev: f64, ) -> Result<DiimfpEvaluation, ElectronDataError>

Self::diimfp_per_m_ev with the number of loss-function evaluations of its momentum integral whose own ω_p integral did not converge (#298, #368); see DiimfpEvaluation. The value is the same number Self::diimfp_per_m_ev returns, and the errors are the same.

pub fn imfp_unconverged( &self, energy_ev: f64, ) -> Result<(u64, f64), ElectronDataError>

Cells of the outer ω_p integral of Self::imfp_and_stopping at energy_ev that were accepted without meeting their tolerance (budget of the bisection spent, maximum depth, or a midpoint that is not representable; #298), and the error estimate of those cells relative to λ⁻¹. (0, 0.0) when the integral converged.

pub fn imfp_and_stopping( &self, energy_ev: f64, ) -> Result<InelasticPoint, ElectronDataError>

The inverse IMFP (m⁻¹) and the stopping power (eV/m) at kinetic energy energy_ev above the Fermi level.

pub fn imfp_and_stopping_evaluation( &self, energy_ev: f64, ) -> Result<ImfpEvaluation, ElectronDataError>

Self::imfp_and_stopping with the cells of its outer ω_p integral that were accepted without meeting their tolerance (#298, #368), from the same single evaluation; see ImfpEvaluation. The point is the same value Self::imfp_and_stopping returns, and the errors are the same.

pub fn imfp_m(&self, energy_ev: f64) -> Result<f64, ElectronDataError>

The inelastic mean free path λ(E), m.

pub fn stopping_power_ev_per_m( &self, energy_ev: f64, ) -> Result<f64, ElectronDataError>

The stopping power S(E), eV/m.

pub fn tabulate( &self, energies_ev: &[f64], ) -> Result<Vec<InelasticPoint>, ElectronDataError>

IMFP and stopping power on a grid of energies, in order. Sequential and deterministic.

pub fn diimfp_grid( &self, energies_ev: &[f64], ) -> Result<DiimfpGrid, ElectronDataError>

The DIIMFP of this model tabulated once for the rows of one inelastic table, whose energies (eV, above the Fermi level) are energies_ev. See DiimfpGrid for the method and its accuracy. The profiles are built in parallel (rayon) and collected in a fixed order, so the grid is bit-identical on any thread count.

Trait Implementations

impl Clone for FullPenn

fn clone(&self) -> FullPenn

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for FullPenn

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.