Empirical computations

JenaAtomicCalculator.Empirical.AbstractEmpiricalApproximation — Type

abstract type Empirical.AbstractEmpiricalApproximation ... defines an abstract and a number of singleton types to distinguish the empirical approximations that are applied for estimating cross sections, rates and (plasma) rate coefficients within this module. Which approximations are supported by a given function follows from its methods (multiple dispatch); an unsupported combination raises an informative error.

+ ScaledHydrogenic     ... hydrogenic (Kramers-type) formulas, scaled by tabulated binding energies;
                           cf. Empirical.scaledBindingEnergy.
+ UsingJAC             ... simple mean-field computations of energies and one-electron amplitudes, replacing a
                           tabulated/scaled input of the SAME formula by a real one: DFS orbitals and E1
                           amplitudes for photoionization/photorecombination/photoemission, cf.
                           Empirical.photoionizationCrossSection; AL-Field orbital energies (the binding energy
                           in the Koopmans sense, cf. Basics.perform(::Empirical.Computation)) for the
                           electron-impact-ionization/three-body-recombination binding energy P, cf.
                           Empirical.impactIonizationBindingEnergy.
+ GivenEinsteinA       ... an externally given Einstein-A value (multipole, energy, rate); cf. below.
+ Lotz1967             ... Lotz's (1967) formula for electron-impact ionization cross sections
                           [W. Lotz, Z. Physik 206, 205 (1967)].
+ VanRegemorter1962    ... Van Regemorter's (1962) formula for electron-impact excitation of optically
                           allowed (E1) transitions [H. Van Regemorter, ApJ 136, 906 (1962)]; selects the
                           formula in Empirical.impactExcitationCrossSection/PlasmaAlpha, which take a
                           separate aSource::AbstractEmpiricalApproximation keyword (ScaledHydrogenic() or
                           GivenEinsteinA(..)) to fix how the underlying Einstein-A value is estimated.
+ ConstantCollisionStrength ... a constant, energy-independent collision strength Omega (of order unity by
                           convention) for electron-impact excitation of an optically forbidden (M1 or E2)
                           transition [M. J. Seaton, Proc. Phys. Soc. 79, 1105 (1962); the Omega ~ 1 convention
                           is also used, e.g., by C. J. Shingles et al., MNRAS 492, 2029 (2020)]; selects the
                           formula in Empirical.forbiddenExcitationCrossSection/PlasmaAlpha. E1 transitions are
                           rejected (use VanRegemorter1962 instead); E3 and higher multipoles are not supported.
+ Bohr1913             ... Bohr's classical stopping power of a plasma, with the Coulomb logarithm
                           ln(1.123 m v^3 / (e^2 omega_p))  [N. Bohr, Philos. Mag. 25, 10 (1913)].
+ Bethe1931            ... Bethe's quantum stopping power of a plasma, with the Coulomb logarithm
                           ln(2 E / (hbar omega_p)) for an electron projectile
                           [H. Bethe, Ann. Phys. (Leipzig) 397, 325 (1930)].
+ KozmaFranson1992     ... the piecewise electron loss function of Kozma & Fransson, quantal above and
                           classical below E = 14 eV [C. Kozma & C. Fransson, ApJ 390, 602 (1992), Eqs. (1-3);
                           after R. Schunk & P. Hays, Planet. Space Sci. 19, 113 (1971)].
+ Axelrod1980          ... Axelrod's relativistic plasma energy loss, i.e. the Bethe-type loss with the plasma
                           energy hbar omega_p as effective ionization potential [T. Axelrod, PhD thesis, UC
                           Santa Cruz (1980); as transcribed by P. Milne et al., ApJS 124, 503 (1999), Eqs. (1,3)].
+ ADK1986              ... the (quasiclassical) tunneling ionization rate of Ammosov, Delone & Krainov for an
                           atom or ion in a strong, quasi-static electric field [M. Ammosov, N. Delone &
                           V. Krainov, Sov. Phys. JETP 64, 1191 (1986); as summarized, e.g., by J. Bauer &
                           P. Mulser, Phys. Rev. A 59, 569 (1999) [arXiv:physics/9802042], Eq. (10)].
+ OverBarrierModel1980 ... the classical (geometric) over-barrier model for single-electron capture in a slow
                           ion-atom/ion-ion collision [N. Bohr & J. Lindhard, K. Dan. Vidensk. Selsk. Mat.-Fys.
                           Medd. 28, No. 7 (1954); cast into cross-section form by H. Ryufuku & T. Watanabe,
                           Phys. Rev. A 19, 1538 (1979); ibid. 20, 1828 (1979)]; selects the formula in
                           Empirical.chargeExchangeCrossSection/PlasmaAlpha. Valid only for slow (adiabatic)
                           collisions, relative velocity v << 1 a.u.; energy-independent within that regime.
+ NiehausScaling1986   ... the empirical q/I_p^2 scaling of single-electron-capture cross sections for slow,
                           highly charged ion-atom/ion-ion collisions [established experimentally, e.g. by
                           M. Kimura et al., J. Phys. B 28, L643 (1995); K. Hosaka et al., Fus. Eng. Design
                           34-35, 781 (1997); shown to be the high-q limit of an extended classical
                           over-barrier model by A. Niehaus, J. Phys. B 19, 2925 (1986)]; selects the formula
                           in Empirical.chargeExchangeCrossSection/PlasmaAlpha. Same regime as
                           OverBarrierModel1980, but grows only as q (not q^2), matching experiment better at
                           high charge states.
+ BelyaevYakovleva2017 ... the simplified Landau-Zener model for inelastic (neutralization, ion-pair
                           formation, excitation, de-excitation) rate coefficients in low-energy A^Z+/A^(Z+1)+
                           collisions with H/H^-, important for non-LTE stellar-atmosphere modeling
                           [A. K. Belyaev & S. A. Yakovleva, A&A 606, A147 (2017) for Z=0;
                           A. K. Belyaev & S. A. Yakovleva, A&A 608, A33 (2017) for Z=1,2]; selects the
                           formulas in Empirical.neutralizationCrossSection/ReducedRate and
                           Empirical.deExcitationCrossSection/ReducedRate. Needs only the electronic bound
                           (ionization) energies and statistical weights of the atomic/ionic states involved,
                           not a full electronic-structure calculation.
+ Arnaud1985DR         ... the semi-empirical dielectronic recombination (DR) rate coefficients of M. Arnaud &
                           R. Rothenflug, A&AS 60, 425 (1985); selects the formula in
                           Empirical.dielectronicRecombinationPlasmaAlpha. Supported only for the H-like
                           sequence (4 tabulated ions: O VIII, Mg XII, Ca XX, Fe XXVI, via
                           PeriodicTable.dielectronicRecombinationParameters_Arnaud1985) and the He-like
                           sequence (a general Z-scaling closed form, any Z); an unsupported isoelectronic
                           sequence raises an informative error rather than silently returning zero.
+ Arnaud1985EA         ... the semi-empirical excitation-autoionization (E-A) contribution to the electron-
                           impact ionization rate, M. Arnaud & R. Rothenflug, A&AS 60, 425 (1985), Appendix A;
                           selects the formula in Empirical.excitationAutoionizationPlasmaAlpha. This is an
                           ADDITIVE contribution -- it must be summed with the existing, unmodified
                           Empirical.impactIonizationPlasmaAlpha (Lotz, direct ionization) to obtain the
                           total EII rate; it neither replaces nor alters that function. Supported only for
                           the Li-like, Na-like and Mg-through-S-like sequences (general, Z-scaling closed
                           forms; a few Z sub-ranges excluded, e.g. Z=17 for Na-like) and 4 explicit "other
                           species" fits (Ca, Ca+, Fe3+, Fe4+); an unsupported combination raises an
                           informative error rather than silently returning zero.
source
JenaAtomicCalculator.Empirical.AbstractStoppingMaterial — Type

abstract type Empirical.AbstractStoppingMaterial ... defines an abstract and a number of types for the material in which the projectile is stopped. In the (Bohr/Bethe-type) stopping formulas of this module, a material enters only through the density of the electrons that take up the energy and through one characteristic energy of these electrons: the plasma energy hbar omegap = hbar sqrt(4 pi ne e^2/m_e) for free electrons, and the mean excitation energy Ibar for bound electrons. Nuclear stopping and radiative (bremsstrahlung) losses belong to different mechanisms and are not included in any of these materials.

+ FreeElectronGas      ... the free electrons of a (fully ionized) plasma with electron density ne.
+ NeutralAtomGas       ... the bound electrons of a neutral, monoatomic gas of atoms (Z, A) with number
                           density natom; the mean excitation energy is estimated as
                           Ibar = 9.1 Z (1 + 1.9 Z^(-2/3)) eV (Segre 1977; Roy & Reed 1968).
+ PartiallyIonizedGas  ... a monoatomic gas of atoms (Z, A) with number density natom, of which chie electrons
                           per atom are ionized: (Z - chie) natom bound electrons with Ibar plus chie natom
                           free electrons with hbar omega_p; cf. Milne et al. (1999, Eq. 3).
source
JenaAtomicCalculator.Empirical.AbstractStoppingProjectile — Type

abstract type Empirical.AbstractStoppingProjectile ... defines an abstract and a number of types for the projectile whose stopping power is requested. The projectile enters the stopping formulas only through its charge z (as z^2 in the prefactor), its mass M (via the velocity v = sqrt(2 E/M) for a given kinetic energy E) and its quantum statistics: an electron projectile is indistinguishable from the target electrons, which limits the maximum energy transfer to E/2 and reduces the Coulomb logarithm by ln 2 relative to a heavy projectile of the same velocity.

+ ElectronProjectile   ... an electron (z = -1, M = 1).
+ PositronProjectile   ... a positron (z = +1, M = 1); it is distinguishable from the target electrons but is
                           treated with the same Coulomb logarithms as an electron -- the differences appear
                           in the relativistic corrections and are beyond the accuracy of these estimates
                           (cf. Milne et al. 1999, who sum electron and positron losses in just this way).
+ IonProjectile        ... a bare (structureless) ion of charge z and mass M [in units of m_e]; for a proton,
                           IonProjectile(1.0, Defaults.PROTON_MASS_U/Defaults.ELECTRON_MASS_U). Electron capture
                           and loss of a dressed ion are not accounted for.
source
JenaAtomicCalculator.Empirical.Computation — Type

struct Empirical.Computation ... defines a type for an empirical computation of various ionization and charge exchange processes.

+ name                   ::String                           ... A name associated to the computation.
+ nuclearModel           ::Nuclear.Model                    ... Model, charge and parameters of the nucleus.
+ grid                   ::Radial.Grid                      ... The radial grid to be used for the computation.
+ configs                ::Array{Configuration,1}           ... A list of non-relativistic configurations.
+ settings               ::Basics.AbstractEmpiricalSettings ... Provides the settings for the selected computations.
source
JenaAtomicCalculator.Empirical.Computation — Method

Empirical.Computation(comp::Empirical.Computation; name::Union{Nothing,String}=nothing, nuclearModel::Union{Nothing,Nuclear.Model}=nothing, grid::Union{Nothing,Radial.Grid}=nothing, configs::Union{Nothing,Array{Configuration,1}}=nothing, settings::Union{Nothing,Any}=nothing, printout::Bool=false)

... constructor for modifying the given Empirical.Computation by 'overwriting' the previously selected parameters.
    A `computation::Empirical.Computation` is returned.
source
JenaAtomicCalculator.Empirical.ConstantCollisionStrength — Type

struct ConstantCollisionStrength <: Empirical.AbstractEmpiricalApproximation ... to select the constant-collision-strength treatment of electron-impact excitation for an optically forbidden (M1 or E2) transition, cf. Empirical.forbiddenExcitationCrossSection/PlasmaAlpha; Omega is the (energy-independent) collision strength, by convention of order unity when no detailed calculation is available [M. J. Seaton, Proc. Phys. Soc. 79, 1105 (1962)].

+ Omega       ::Float64         ... Collision strength (dimensionless); defaults to 1.0.
source
JenaAtomicCalculator.Empirical.GivenEinsteinA — Type

struct GivenEinsteinA <: Empirical.AbstractEmpiricalApproximation ... to communicate an externally given Einstein-A value, together with the multipole and energy of the transition i –> f.

+ multipole   ::EmMultipole     ... Multipole of the transition.
+ energy      ::Float64         ... Energy of transition.
+ rate        ::Float64         ... Einstein-A value.
source
JenaAtomicCalculator.Empirical.InelasticHChannel — Type

struct Empirical.InelasticHChannel ... a single atomic/ionic scattering channel A^Z+(j) + H that correlates to the ionic ground molecular state A^(Z+1)+ + H^- (Belyaev & Yakovleva 2017, Steps 1-2); typically built via Empirical.statisticalWeight for the p_stat field, or – for a new reaction specified via Configuration – automatically by Empirical.InelasticHReaction below.

+ name        ::String    ... a short label for the channel, e.g. "Ba+(6d 2D)", for display purposes only.
+ E           ::Float64   ... electronic bound energy Ej [a.u.], negative, measured from the ionization limit.
+ pstat       ::Float64   ... statistical probability p_stat_j of this channel within the relevant molecular
                              symmetry (Eqs. 1-2); e.g. taken from Table 1 of Belyaev & Yakovleva (2017) for a
                              known case, or derived from the relevant molecular-term-symmetry correlation.
source
JenaAtomicCalculator.Empirical.InelasticHReaction — Type

struct Empirical.InelasticHReaction ... specifies an inelastic ion + hydrogen reaction A^(Z+1)+ + H^- <-> A^Z+(f) + H directly in terms of JAC's own Nuclear.Model and Configuration entities, for one or several final configurations at once.

+ nm        ::Nuclear.Model            ... the projectile ion's nuclear model; only nm.Z is used here (for the covalent
                                            species' charge and for looking up its standard atomic mass).
+ iConfIon  ::Configuration             ... the (closed-shell) ground configuration of the ionic entrance species A^(Z+1)+.
+ iConfH    ::Configuration             ... either H^-(1s^2) or H(1s^1) -- this module only ever considers hydrogen.
+ fConfIon  ::Array{Configuration,1}    ... the final ion configurations A^Z+(f) of interest; each must differ from iConfIon
                                            by exactly one electron in exactly one shell.
+ fConfH    ::Configuration             ... the complementary final hydrogen configuration; derived automatically from
                                            iConfH by electron-count conservation if not supplied explicitly.
source
JenaAtomicCalculator.Empirical.InelasticHReaction — Method

Empirical.InelasticHReaction(nm::Nuclear.Model, iConfIon::Configuration, iConfH::Configuration, fConfIon::Array{Configuration,1}; fConfH::Union{Nothing,Configuration}=nothing) ... outer constructor that validates iConfH/fConfH and derives fConfH from iConfH by electron-count conservation if not given explicitly. Raises an informative error if iConfH is not H(1s^1) or H^-(1s^2), or if a given fConfH is inconsistent with iConfH's electron count. An Empirical.InelasticHReaction is returned.

source
JenaAtomicCalculator.Empirical.IonProjectile — Type

struct Empirical.IonProjectile <: AbstractStoppingProjectile ... a bare (structureless) ion as projectile of a stopping-power calculation.

+ z           ::Float64   ... charge of the ion [in units of e].
+ M           ::Float64   ... mass of the ion [in units of m_e].
source
JenaAtomicCalculator.Empirical.NeutralAtomGas — Type

struct Empirical.NeutralAtomGas <: AbstractStoppingMaterial ... the bound electrons of a neutral, monoatomic gas.

+ Z           ::Int64     ... nuclear charge of the atoms.
+ A           ::Float64   ... atomic mass number of the atoms; only used to express the mass stopping power.
+ natom       ::Float64   ... atom density [a.u.].
source
JenaAtomicCalculator.Empirical.PartiallyIonizedGas — Type

struct Empirical.PartiallyIonizedGas <: AbstractStoppingMaterial ... a partially ionized, monoatomic gas: (Z - chie) bound electrons per atom plus chie free electrons per atom.

+ Z           ::Int64     ... nuclear charge of the atoms.
+ A           ::Float64   ... atomic mass number of the atoms; only used to express the mass stopping power.
+ chie        ::Float64   ... number of free electrons per atom (ionization fraction), 0 <= chie <= Z.
+ natom       ::Float64   ... atom density [a.u.].
source
JenaAtomicCalculator.Basics.perform — Method

Basics.perform(computation::Empirical.Computation; output::Bool=false) ... to set-up and perform an empirical computation that starts from a given nuclear model and set of configurations, and which is mainly controlled by its settings. The results are printed to screen. With output = false, nothing is returned. With output = true, a results::Basics.PerformResults is returned, an AbstractDict{String,Any} whose keys depend on the kind and the specifications of the empirical computation.

source
JenaAtomicCalculator.Empirical.activeShell — Method

Empirical.activeShell(iConf::Configuration, fConf::Configuration) ... to identify the single shell whose occupation differs between iConf and fConf, and check that it differs by exactly one electron (a genuine one-electron transfer) – the scope Empirical.InelasticHReaction is restricted to (see the section note above). Raises an informative error if more than one shell differs, or if the occupation change is not exactly one electron. A sh::Shell is returned.

source
JenaAtomicCalculator.Empirical.adkAngularFactor — Method

Empirical.adkAngularFactor(l::Int64, m::Int64) ... to provide the angular factor f(l,m) = (2l+1) (l+|m|)! / ( 2^|m| |m|! (l-|m|)! ) of the ADK rate, which depends on the actual orbital and magnetic quantum numbers l and m of the initial (bound) state – in contrast to the coefficient C^2_(nl), which uses the effective quantum numbers n* and l*. A wa::Float64 is returned.

source
JenaAtomicCalculator.Empirical.adkCoefficient — Method

Empirical.adkCoefficient(nstar::Float64, lstar::Float64) ... to provide the squared ADK coefficient C^2(n* l) = 2^(2n) / ( n* Gamma(n+l+1) Gamma(n-l) ) , using the Gamma function so that n* and l* need not be integers, as appropriate for a real atom or ion (only for a pure hydrogenic 1s state, Z = n* = 1 and l* = 0, are they integers by construction). For this hydrogenic case, C^2(1,0) = 4 exactly, and the full ADK rate below then reduces exactly to the classical Landau tunneling formula for hydrogen [Bauer & Mulser, Eq. (8)] – a useful, and here numerically verified, cross check of the normalization. Many practical implementations instead use the quasiclassical (Stirling) approximation C^2_(n) = (2e/n)^(2n)/(2 pi n), which drops the l-dependence altogether and is accurate only for n >> 1; for n* = 1 it overestimates C^2 by 17.6%, and the Gamma-function form is used throughout this module instead. A c2::Float64 is returned.

source
JenaAtomicCalculator.Empirical.augerRate — Method

Empirical.augerRate(iConf::Configuration, fConf::Configuration, approx::Empirical.ScaledHydrogenic; printout::Bool=false, data::PeriodicTable.AbstractEnergyData=PeriodicTable.XrayDataBooklet(), yieldData::PeriodicTable.AbstractYieldData=PeriodicTable.KrauseAdopted2016()) ... to estimate empirically the (total) K-shell Auger rate Aauger for filling the K-shell vacancy of the transition iConf -> fConf, from the fluorescence yield omegaK and the radiative K-shell rate. The two decay channels of a K-shell hole share the same total rate Gamma = Arad + Aauger, and their branching is fixed by the yield, Arad = omegaK Gamma, so that Aauger = Arad (1 - omegaK) / omegaK, where A_rad is the (empirical) radiative rate of iConf -> fConf, cf. Empirical.photoemissionEinsteinA. A rate::Float64 [a.u.] is returned. Quantity: a spontaneous rate [1/s] – an intrinsic property of the ion, independent of the plasma.

    Note: This estimate exploits that K-shell Auger rates are nearly independent of Z along an isoelectronic
          sequence, whereas radiative rates scale approximately as Z^4; the whole Z dependence is therefore
          carried by omega_K. It presumes that iConf -> fConf refills a *K-shell* (1s) vacancy; omega_K is taken
          for the neutral element and is only a rough guide for an ion.

    Note: The branching ratio A_auger/A_rad = (1 - omega_K)/omega_K is accurate (it is the definition of the
          yield), but the *absolute* Auger rate is only as good as the radiative rate A_rad that feeds it. In the
          ScaledHydrogenic approximation A_rad is itself good to about a factor 1.5, and this uncertainty is
          amplified by the large factor (1 - omega_K)/omega_K for light elements: for neutral Ne the implied
          K-shell width A_rad + A_auger comes out near 2.6 eV, against a measured natural width of ~0.27 eV.
          The result should therefore be read as an order-of-magnitude estimate, most reliable near omega_K ~ 0.5.
source
JenaAtomicCalculator.Empirical.augerYield — Method

Empirical.augerYield(Z::Int64; data::PeriodicTable.AbstractYieldData=PeriodicTable.KrauseAdopted2016()) ... to provide the K-shell Auger yield aK = 1 - omegaK of the neutral element with nuclear charge Z, i.e. the probability that a K-shell vacancy decays by a (radiationless) Auger transition. A yield::Float64 in [0,1] is returned.

source
JenaAtomicCalculator.Empirical.bindingEnergy — Method

Empirical.bindingEnergy(Z::Int64, sh::Shell; data::PeriodicTable.AbstractEnergyData=XrayDataBooklet() ) ... to estimate the binding) energy of the given shell for a neutral atom; these binding energies are taken from semi-empirical tabulations due to data by different data sets. An energy::Float64 > 0. is returned. ...... Perhaps, introduce an optional argument scaleLast::Bool to scale values, if no explicit binding energies are available.

source
JenaAtomicCalculator.Empirical.bindingEnergy — Method

Empirical.bindingEnergy(Z::Int64, subsh::Subshell, conf::Configuration; data::PeriodicTable.AbstractEnergyData=XrayDataBooklet()) ... to evaluate the binding energy of a subshell electron if part of the given configuration. The tabulated binding energies refer to the neutral atom; this procedure assumes that each missing electron (compared with the neutral system) adds 0.3 / n Hartree to the binding energy, where n is the principal quantum number of the subshell. It also requires that the shell associated with subsh is occupied in the given configuration. An energy::Float64 is returned.

source
JenaAtomicCalculator.Empirical.bindingEnergy — Method

Empirical.bindingEnergy(Z::Int64, subsh::Subshell; data::PeriodicTable.AbstractEnergyData=XrayDataBooklet() ) ... to estimate the binding) energy of the given subshell for a neutral atom; these binding energies are taken from semi-empirical tabulations due to data by different data sets. An energy::Float64 > 0. is returned.

source
JenaAtomicCalculator.Empirical.boundFreeGauntFactor — Method

Empirical.boundFreeGauntFactor(n::Int64, x::Float64) ... to provide the bound-free Gaunt factor g^(bf) for the photoionization of shell n at the photon energy x = omega/omegathreshold, i.e. the quantum correction to Kramers' semiclassical cross section, sigma = sigma^(Kramers) * g^(bf). For n = 1 the factor is exact (nonrelativistic): it is the ratio of Stobbe's (1930) closed-form 1s cross section, sigma^(Stobbe)(x) = 2^9 pi^2/(3 e^4) alpha ao^2 x^(-4) exp(4 - 4 arctan(kappa)/kappa) / (1 - e^(-2 pi/kappa)) with kappa = sqrt(x-1), to the Kramers form: at threshold, sigma^(Stobbe) = 6.304 Mb (Z = 1) analytically, so that g(1) = 6.304/7.907 = 0.7973, while at large x the exact cross section falls off as x^(-7/2) against Kramers' x^(-3), so that g ~ x^(-1/2). For n >= 2 the deviation from unity is scaled with the leading-order n-dependence of the Menzel & Pekeris (1935) expansion, g = 1 + [g_1s(x) - 1] n^(-2/3), which preserves the exact n = 1 limit and approaches the correct semiclassical limit g -> 1 for large n, where Kramers' formula becomes exact. A gbf::Float64 is returned; g = 1 is returned for x <= 1, where no cross section remains to be corrected.

source
JenaAtomicCalculator.Empirical.chargeExchangeCaptureShell — Method

Empirical.chargeExchangeCaptureShell(q::Float64, Ip::Float64) ... to estimate the principal quantum number of the shell into which the active electron is most likely captured in a slow-collision charge-exchange event, by the standard hydrogenic level-matching argument associated with the over-barrier model: the electron is assumed to be captured near the point where the H-like binding energy of shell n in the acceptor ion of charge q, En = q^2/(2 n^2), matches the donor's own ionization potential Ip [a.u.], i.e. nc = q / sqrt(2 Ip) . This level-matching argument is commonly invoked alongside the over-barrier model [e.g. H. Ryufuku, K. Sasaki & T. Watanabe, Phys. Rev. A 21, 745 (1980); A. Niehaus, J. Phys. B 19, 2925 (1986)] to identify the dominant shell of capture; it is a direct algebraic consequence of the hydrogenic formula already used throughout this module (cf. Empirical.scaledBindingEnergy), not an independently fitted quantity. A named tuple (nc::Float64=, n::Int64=) is returned: nc is the exact (generally non-integer) level-matching value, and n = max(1, round(Int64, nc)) is the nearest physical shell. Quantity: n and nc are dimensionless (quantum) labels, not an energy or a cross section.

    Note: this identifies only the single *dominant* shell; the true capture probability is known to spread
          over a range of a few adjacent n (particularly when nc falls near a half-integer), which this
          estimate does not model. Empirical.chargeExchangeStateSelectiveCrossSection() assumes *all* capture
          goes into this one shell -- a further, explicit simplification beyond the total-cross-section
          estimate itself.
source
JenaAtomicCalculator.Empirical.chargeExchangeCrossSection — Method

Empirical.chargeExchangeCrossSection(q::Float64, targetConf::Configuration, targetZ::Float64, approx::Empirical.NiehausScaling1986; printout::Bool=false, velocity::Union{Nothing,Float64}=nothing) ... convenience wrapper that determines the donor's ionization potential from targetConf via Empirical.ionizationPotential(targetZ, targetConf) – valid for any standard-filling configuration, neutral atom or ion alike, so the same call handles both atom-ion and ion-ion charge exchange – and dispatches to the direct (q, Ip) method above. A sigma::Float64 [a.u.] is returned.

source
JenaAtomicCalculator.Empirical.chargeExchangeCrossSection — Method

Empirical.chargeExchangeCrossSection(q::Float64, targetConf::Configuration, targetZ::Float64, approx::Empirical.OverBarrierModel1980; printout::Bool=false, velocity::Union{Nothing,Float64}=nothing) ... convenience wrapper that determines the donor's ionization potential from targetConf via Empirical.ionizationPotential(targetZ, targetConf) – valid for any standard-filling configuration, neutral atom or ion alike, so the same call handles both atom-ion and ion-ion charge exchange – and dispatches to the direct (q, Ip) method above. A sigma::Float64 [a.u.] is returned.

source
JenaAtomicCalculator.Empirical.chargeExchangeCrossSection — Method

Empirical.chargeExchangeCrossSection(q::Float64, Ip::Float64, approx::Empirical.NiehausScaling1986; printout::Bool=false, velocity::Union{Nothing,Float64}=nothing) ... to estimate the total single-electron-capture cross section for a bare (or effectively bare) ion of charge q colliding slowly with a donor of ionization potential Ip [a.u.], by the empirical high-charge scaling law sigma = 2.6e-13 * q / Ip[eV]^2 [cm^2] , established experimentally for slow (v < 1 a.u.) collisions of highly charged ions with atoms [e.g. M. Kimura et al., J. Phys. B 28, L643 (1995); K. Hosaka et al., Fus. Eng. Design 34-35, 781 (1997)], and shown to be the high-q limit of an extended classical over-barrier model [A. Niehaus, J. Phys. B 19, 2925 (1986)]. A sigma::Float64 [a.u.] is returned. Quantity: a cross section [a.u.] – energy-independent within the slow-collision regime (see the module note above); fold with a relative-velocity distribution, or pass to Empirical.chargeExchangePlasmaAlpha, to obtain a rate coefficient.

    Note: this scaling grows only linearly in q, matching experiment better than the plain (q^2-growing)
          Empirical.OverBarrierModel1980 at high charge states; it is a global fit and not expected to be
          accurate for individual, near-resonant systems (where the specific energy defect of the reaction
          matters), nor does it provide state (n,l) selectivity or multi-electron capture.
    Note: if velocity (relative collision velocity [a.u.]) is given, a warning is issued once it approaches or
          exceeds the adiabatic regime, v ~ 1 a.u.
source
JenaAtomicCalculator.Empirical.chargeExchangeCrossSection — Method

Empirical.chargeExchangeCrossSection(q::Float64, Ip::Float64, approx::Empirical.OverBarrierModel1980; printout::Bool=false, velocity::Union{Nothing,Float64}=nothing) ... to estimate the total single-electron-capture cross section for a bare (or effectively bare) ion of charge q colliding slowly with a donor of ionization potential Ip [a.u.], by the classical (geometric) over-barrier model of Bohr & Lindhard, cast into cross-section form by Ryufuku & Watanabe [H. Ryufuku & T. Watanabe, Phys. Rev. A 19, 1538 (1979); ibid. 20, 1828 (1979)]: Rc = (q + 1) / (2 Ip), sigma = pi Rc^2 , with R_c the critical (saddle-point) internuclear distance at which the potential barrier between the two Coulomb centers drops below the donor's binding energy, allowing the active electron to move classically onto the (empty) projectile. A sigma::Float64 [a.u.] is returned. Quantity: a cross section [a.u.] – energy-independent within the slow-collision regime (see the module note above); fold with a relative-velocity distribution, or pass to Empirical.chargeExchangePlasmaAlpha, to obtain a rate coefficient.

    Note: this is the single-active-electron, energy-independent (adiabatic) limit; its q^2 growth at high
          charge states is known to *overestimate* the cross section there -- cf. Empirical.NiehausScaling1986,
          whose q^1 growth matches measurements better for highly charged ions. Good to a factor of a few for
          low/moderate q; no state (n,l) selectivity or multi-electron capture is provided.
    Note: if velocity (relative collision velocity [a.u.]) is given, a warning is issued once it approaches or
          exceeds the adiabatic regime, v ~ 1 a.u.
source
JenaAtomicCalculator.Empirical.chargeExchangePlasmaAlpha — Method

Empirical.chargeExchangePlasmaAlpha(T::Float64, q::Float64, targetConf::Configuration, targetZ::Float64, Mproj::Float64, Mtarget::Float64, approx::Empirical.AbstractEmpiricalApproximation; printout::Bool=false) ... convenience wrapper that determines the donor's ionization potential from targetConf via Empirical.ionizationPotential(targetZ, targetConf) – valid for any standard-filling configuration, neutral atom or ion alike – and dispatches to the direct (T, q, Ip, Mproj, Mtarget) method above. An alpha::Float64 [a.u.] is returned.

source
JenaAtomicCalculator.Empirical.chargeExchangePlasmaAlpha — Method

Empirical.chargeExchangePlasmaAlpha(T::Float64, q::Float64, Ip::Float64, Mproj::Float64, Mtarget::Float64, approx::Empirical.AbstractEmpiricalApproximation; printout::Bool=false) ... to estimate the charge-exchange plasma rate coefficient alpha^(CX) for a Maxwellian relative-velocity distribution at temperature T [a.u.], by folding the (velocity-independent, within the slow-collision regime) cross section with the mean relative speed of a Maxwell-Boltzmann distribution, alpha^(CX) (T) = <v> sigma = sqrt(8 T / (pi mu)) * sigma , with mu = Mproj Mtarget / (Mproj + Mtarget) [me] the reduced mass of the collision partners (in the same mass unit as, e.g., Empirical.IonProjectile: for a proton, Defaults.PROTONMASSU/Defaults.ELECTRONMASSU). Unlike the electron-impact plasma rate coefficients elsewhere in this module, no Gauss-Legendre quadrature is needed here: sigma is constant within the slow-collision regime (see the module note above), so <v sigma> = <v> sigma analytically. approx selects the cross-section formula (OverBarrierModel1980 or NiehausScaling1986) via Empirical.chargeExchangeCrossSection; any other approx raises the same informative MethodError as an unsupported cross-section call. An alpha::Float64 [a.u.] is returned. Quantity: a rate coefficient [cm^3/s] – multiply by the target (donor) number density nB [1/cm^3] to obtain the rate per (projectile) ion [1/s].

    Note: a warning is issued if the thermal relative speed sqrt(2 T/mu) approaches or exceeds the adiabatic
          regime (v ~< 1 a.u.), since the underlying cross section becomes unreliable for a growing fraction
          of the distribution.
source
JenaAtomicCalculator.Empirical.chargeExchangePrintout — Method

Empirical.chargeExchangePrintout(sa::String, q::Float64, Ip::Float64, velocity::Union{Nothing,Float64}, sigma::Float64) ... to print a common report about a charge-exchange cross-section estimate: the approximation and its regime of validity (sa), the projectile charge and donor ionization potential, the relative velocity if given, and the resulting cross section in the user-defined unit. Nothing is returned.

source
JenaAtomicCalculator.Empirical.chargeExchangeStateSelectiveCrossSection — Method

Empirical.chargeExchangeStateSelectiveCrossSection(q::Float64, targetConf::Configuration, targetZ::Float64, approx::Empirical.AbstractEmpiricalApproximation; printout::Bool=false) ... convenience wrapper that determines the donor's ionization potential from targetConf via Empirical.ionizationPotential(targetZ, targetConf) – valid for any standard-filling configuration, neutral atom or ion alike – and dispatches to the direct (q, Ip, approx) method above. A named tuple (n::Int64=, nc::Float64=, sigmaTotal::Float64=, states::Array=) is returned.

source
JenaAtomicCalculator.Empirical.chargeExchangeStateSelectiveCrossSection — Method

Empirical.chargeExchangeStateSelectiveCrossSection(q::Float64, Ip::Float64, approx::Empirical.AbstractEmpiricalApproximation; printout::Bool=false) ... to split the total single-electron-capture cross section (Empirical.chargeExchangeCrossSection) into a state-selective (n,l) estimate: all capture is assumed to proceed into the single dominant shell n = Empirical.chargeExchangeCaptureShell(q, Ip).n, distributed over its orbital substates l = 0, ..., n-1 by the purely statistical hydrogenic degeneracy weight, sigmal = sigmatotal * (2l+1) / n^2 , with no dynamical (angular-momentum-transfer) preference among them – the simplest defensible assumption absent a detailed (e.g. Landau-Zener or close-coupling) calculation. approx selects the total-cross-section formula (OverBarrierModel1980 or NiehausScaling1986) via Empirical.chargeExchangeCrossSection; any other approx raises the same informative MethodError as an unsupported cross-section call. A named tuple (n::Int64=, nc::Float64=, sigmaTotal::Float64=, states::Array=) is returned, where states is an array of named tuples (l::Int64=, sigma::Float64=, fraction::Float64=) with sum(fraction) == 1 and sum(sigma) == sigmaTotal by construction. Quantity: sigmaTotal and each sigma [a.u.] are cross sections, energy-independent within the slow-collision regime, exactly as Empirical.chargeExchangeCrossSection; fraction is dimensionless.

    Note: the function name is deliberately explicit -- "StateSelective" -- to distinguish it from
          Empirical.chargeExchangeCrossSection, which returns only the total, state-summed cross section. The
          n,l-selectivity here rests on the single-dominant-shell and statistical-l simplifications described
          above and in Empirical.chargeExchangeCaptureShell; it is considerably cruder than the total cross
          section itself and should be read as indicative of the capture-shell order and a plausible
          l-spread, not as a precise state-resolved prediction.
source
JenaAtomicCalculator.Empirical.checkCrossingRadius — Method

Empirical.checkCrossingRadius(Rj::Float64; RjRefuse::Float64=4.5, RjWarn::Float64=10.0, label::String="") ... to decide whether the ionic-covalent crossing at R_j (Empirical.nonadiabaticRadius) lies in the LONG-RANGE region the Belyaev-Yakovleva model is built for. The model assumes a single, isolated, diabatic crossing in the asymptotic region where the two partners' charge clouds do not yet interpenetrate, and the Olson-Smith-Bauer coupling of Eq. (12) is a long-range asymptotic formula; nothing in the arithmetic itself says so, so a deeply bound final state is answered for as confidently as a Rydberg one and returns a large, wholly spurious rate. Three tiers are returned as a Symbol: :ok above RjWarn, :marginal between RjWarn and RjRefuse (computed, with a warning – these are the small rates the paper's own Sec. 2.4 says the model does not target), and :refuse below RjRefuse, where the caller should return a zero rather than a number. A verdict::Symbol is returned.

    THE TWO DEFAULTS ARE MEASURED, NOT CHOSEN, and the measurement is `tools/probeInelasticHCollisionGate.jl`:

      +  FROM BELOW.  For O+ + H^- the three deeply bound terms O(2p^4 ^3P, ^1D, ^1S) sit at R_j = 2.12, 2.50 and
         3.13 a_0, and BOTH published quantum treatments -- Barklem, A&A 610, A57 (2018), asymptotic LCAO plus
         multichannel Landau-Zener, and Belyaev et al., MNRAS 487, 5097 (2019), MRCI plus quantum probability current --
         give EXACTLY ZERO there, while this model gives 4.3e-10, 2.3e-10 and 8.7e-11 cm^3/s.
      +  FROM ABOVE.  The deepest channel of Belyaev & Yakovleva's OWN Table 1, Ba^+(6s) at -10.0080 eV with Z = 1,
         sits at R_j = 5.88 a_0 and is tabulated by the paper itself, so no gate may refuse it.

    RjRefuse = 4.5 a_0 is the midpoint of the resulting window (3.13, 5.88), clear of both ends;  it is also about twice
    the mean radius of H^-, i.e. the separation at which the two charge clouds begin to overlap.  RjWarn = 10.0 a_0 marks
    where the paper's own small-rate channels begin (Ba^+(6s), (5d), (6p) at 5.88, 6.34, 8.24 a_0).
source
JenaAtomicCalculator.Empirical.deExcitationCrossSection — Method

Empirical.deExcitationCrossSection(E::Float64, Ei::Float64, Ef::Float64, EHminus::Float64, Z::Float64, mu::Float64; printout::Bool=false) ... to compute the (three-channel-approximation) cross section for the de-excitation process A^Z+(i) + H -> A^Z+(f) + H at collision energy E [a.u.] (Ei more deeply bound than Ef, i.e. Ei < Ef is NOT required by this function – the caller must ensure Ei corresponds to the higher-lying, i.e. less deeply bound, initial state for a genuine de-excitation), by summing the two-crossing Landau-Zener transition probability P^Dif(J,E) = 2 pf (1-pf) (1-pi) (Eq. 17) over total angular momentum J, sigma^Dif(E) = pi / (2 mu E) * sumJ (2J+1) P^D_if(J,E) , Eq. (16) of Belyaev & Yakovleva, A&A 608, A33 (2017), with the radial velocities at the two crossings from Eqs. (18)-(19). A sigma::Float64 [a.u.] is returned. Quantity: a cross section [a.u.] – fold with a Maxwellian velocity distribution, or use Empirical.deExcitationReducedRate, to obtain a (reduced) rate coefficient.

    OPEN ISSUE, PAUSED (not resolved -- deliberately set aside, see project_inelastic_h_collisions.md for the
    full trail): checked against Belyaev & Yakovleva (2017), Table 2, for the Ba+(7p) -> Ba+(6d) de-excitation
    rate coefficient at T=6000 K, this formula (after multiplying the reduced rate by p_stat_i) gives
    ~4.6e-10 cm^3/s vs. the paper's quoted 8.94e-10 cm^3/s -- a factor ~1.9 too small, growing to ~2.6 for
    larger energy defects (checked across all 6 de-excitation pairs from the Ba+(7p) row of Table 2). Neither
    dropping the (1-p_i) factor entirely nor removing the Jmax_i restriction resolves this (tested explicitly).
    The likely ROOT CAUSE, identified via Belyaev, Phys. Rev. A 48, 4299 (1993)
    [examples/papers/a93.pra-belyaev-charge-exchange.pdf]: this function treats de-excitation as an isolated
    3-state (i, ionic, f) problem, ignoring that the ionic curve also crosses the OTHER covalent channels of
    the same ion along the way -- that paper's Eq. (3.8) (a general multichannel Landau-Zener branching
    formula) and its accompanying discussion explicitly warn that neglecting such channels can inflate the
    result by a factor of 2 (bounded, for channels beyond the target) or more (unbounded, for channels between
    the initial and target states) -- consistent in both magnitude and trend with what is observed here. A
    properly re-derived attempt at the correct multichannel mapping (25-Jul-2026, see the memory file for the
    full trail) still did NOT succeed -- the correction moved the result FURTHER from the literature value, and
    a sanity check of Eq. (3.8) itself revealed an unresolved inconsistency in even the simplest nontrivial
    case. By contrast, Empirical.neutralizationReducedRate agrees with the paper to 3 significant figures for
    the same system. Treat de-excitation/excitation rates from this function as order-of-magnitude/qualitative-
    trend estimates only; do not rely on them for a quantitative comparison.
source
JenaAtomicCalculator.Empirical.deExcitationReducedRate — Method

Empirical.deExcitationReducedRate(T::Float64, Ei::Float64, Ef::Float64, EHminus::Float64, Z::Float64, mu::Float64; printout::Bool=false, zerosGL::Int64=64, checkRadius::Bool=true) ... to compute the reduced rate coefficient Dif(T; Ei, Ef) for the de-excitation process A^Z+(i) + H -> A^Z+(f) + H at temperature T [a.u.], by folding Empirical.deExcitationCrossSection with the Maxwell distribution, Dif(T; Ei, Ef) = sqrt(8/(pi mu (kT)^3)) int0^infty sigma^Dif(E) E exp(-E/kT) dE , Eq. (15) of Belyaev & Yakovleva, A&A 608, A33 (2017), evaluated by Gauss-Legendre quadrature over [0, 30 kT]. A value::Float64 [a.u.] is returned. Quantity: a (reduced) rate coefficient [cm^3/s] – multiply by the statistical probability pstat of the initial state i, Eq. (2), to obtain the actual rate coefficient Kif(T); cf. Empirical.detailedBalanceRate for the inverse (excitation) process.

source
JenaAtomicCalculator.Empirical.detailedBalanceRate — Method

Empirical.detailedBalanceRate(Kjk::Float64, pstatJ::Float64, pstatK::Float64, deltaEjk::Float64, T::Float64) ... to compute the rate coefficient Kkj(T) for the endothermic inverse (ion-pair formation, or excitation) process k -> j from the rate coefficient Kjk(T) of the exothermic direct (neutralization, or de-excitation) process j -> k, by the detailed balance relation Kkj(T) = Kjk(T) * (pstatk / pstatj) * exp( -deltaEjk / kT ) , Eq. (3) of Belyaev & Yakovleva, A&A 608, A33 (2017), where deltaEjk = Ej - Ek > 0 is the (positive) energy defect of the exothermic j -> k process. A value::Float64 [same unit as Kjk] is returned.

source
JenaAtomicCalculator.Empirical.dielectronicRecombinationPlasmaAlpha — Method

Empirical.dielectronicRecombinationPlasmaAlpha(eDist::Distribution.ElectronMaxwell, iConf::Configuration; approx::Empirical.AbstractEmpiricalApproximation=Arnaud1985DR(), printout::Bool=false) ... to estimate empirically the dielectronic recombination (DR) plasma rate coefficient alpha^(DR)(Te) for the resonant-capture-plus-stabilization channel iConf + e -> (doubly-excited)* -> (radiative decay), following the semi-empirical fits of M. Arnaud & R. Rothenflug, A&AS 60, 425 (1985). Only two isoelectronic sequences of the recombining ion iConf are supported by that paper with self-contained, reproducible parameters (everything else relies on external calculations not tabulated there): + H-like (iConf.NoElectrons == 1): a per-ion fit alphad(T) = ADI T^(-3/2) exp(-T0/T), tabulated for only 4 ions – O VIII, Mg XII, Ca XX, Fe XXVI – via PeriodicTable.dielectronicRecombinationParametersArnaud1985(Z). + He-like (iConf.NoElectrons == 2): a general Z-scaling closed form (any Z >= 3), Sect. 3.2.2 of the paper: alphad(T) = 7.6e-11 A(z) exp(-D(z)/chi) / chi^(3/2) cm^3/s, chi = kT/ILi, z = Z-2, with ILi the ionization potential of the corresponding Li-like ion (taken here from the already-tabulated PeriodicTable.ionizationPotentialsNist2025, i.e. its (Z-2)-th successive ionization potential, rather than re-transcribing further numbers from the 1985 paper). Any other electron number, or approx != Arnaud1985DR(), raises an informative error rather than silently returning zero. An alpha::Float64 [a.u.] is returned. Quantity: a (two-body) recombination rate coefficient [cm^3/s] – multiply by the electron number density ne [1/cm^3] to obtain the rate per ion [1/s].

source
JenaAtomicCalculator.Empirical.effectiveGauntFactor — Method

Empirical.effectiveGauntFactor(x2::Float64, isIon::Bool) ... to provide Van Regemorter's (1962) effective Gaunt factor gbar for electron-impact excitation of optically allowed transitions; x2 = (eps - deltaE)/deltaE is the final electron energy in units of the transition energy. For positive ions, the paper prescribes gbar = 0.2 up to x2 = 2 and the hydrogen 1s -> 2p curve beyond; its high-energy form [Eq. (10)] gbar = sqrt(3)/(2 pi) ln(x2) reaches 0.2 just at x2 = 2.06, so that both branches join seamlessly in gbar = max(0.2, sqrt(3)/(2 pi) ln(x2)). For neutral atoms, the near-threshold behavior [Eq. (9)] gbar = 0.074 x (1 + x2) is applied and capped by the ion branch; the cross section then vanishes at threshold, as it must for a neutral target. A gbar::Float64 is returned.

source
JenaAtomicCalculator.Empirical.effectiveQuantumNumbers — Method

Empirical.effectiveQuantumNumbers(Z::Float64, Ip::Float64) ... to provide the effective principal quantum number n* of an electron with binding energy Ip [a.u.] in the Coulomb tail of an ion of residual charge Z (i.e. the charge seen by the escaping electron once it is far from the core, at asymptotic distances), following the hydrogenic relation Ip = Z^2/(2 n^2), i.e. n = Z / sqrt(2 Ip). A tuple (nstar::Float64, kappa::Float64) is returned, with kappa = sqrt(2 Ip) the (inverse) tunneling decay constant that enters the ADK exponent and the critical (barrier-suppression) field F_c = kappa^3/16.

source
JenaAtomicCalculator.Empirical.electricFieldFromIntensity — Method

Empirical.electricFieldFromIntensity(intensity::Float64) ... to convert a (peak) laser intensity [W/cm^2] to the corresponding (peak) electric field strength [a.u.], via the standard atomic-unit relation intensity[a.u.] = field[a.u.]^2, i.e. F [a.u.] = sqrt( intensity [W/cm^2] / 3.50944758e16 ) . A field::Float64 [a.u.] is returned.

source
JenaAtomicCalculator.Empirical.energyOf — Method

Empirical.energyOf(energies::Array{Pair{Configuration,Float64},1}, conf::Configuration) ... to look up conf's total energy [a.u.] in energies by VALUE equality. The linear search dates from a time when Configuration had a working == but no matching hash, so a Dict{Configuration,Float64} silently dropped valid keys; that bug is fixed (12-Sep-2026) and a Dict would serve now. The list is short and the search is not a cost worth changing a signature for. Raises an informative error if conf is not found. A value::Float64 is returned.

source
JenaAtomicCalculator.Empirical.excitationAutoionizationPlasmaAlpha — Method

Empirical.excitationAutoionizationPlasmaAlpha(eDist::Distribution.ElectronMaxwell, iConf::Configuration; approx::Empirical.AbstractEmpiricalApproximation=Arnaud1985EA(), printout::Bool=false) ... to estimate empirically the excitation-autoionization (E-A) contribution to the total electron-impact ionization plasma rate coefficient of the ion iConf, following the semi-empirical fits of M. Arnaud & R. Rothenflug, A&AS 60, 425 (1985), their Appendix A. This is an ADDITIVE contribution: the paper's own total ionization rate is alpha^(EII: total) = alpha^(EII: direct) + alpha^(EA), so this function returns ONLY the E-A piece and must be summed with Empirical.impactIonizationPlasmaAlpha (the existing, unmodified, Lotz-based direct-ionization rate) to obtain the total – it does not replace or alter that function, nor does it need an fConf (Arnaud & Rothenflug's fits give one aggregate E-A rate per ion, not a shell- or channel-resolved one).

    Supported isoelectronic sequences of iConf (identified by iConf.NoElectrons, with a few further
    (Z, NoElectrons)-specific "other species" cases checked first):
        + Li-like  (NoElectrons ==  3): general Z formula, any Z; explicit correction factors 0.6/0.8/1.25
          for C/N/O.
        + Na-like  (NoElectrons == 11): general Z formula, Z <= 16 or 18 <= Z <= 28 (Z = 17 falls in a gap
          between the paper's two fitted ranges and is not covered).
        + Mg/Al/Si/P/S-like (NoElectrons == 12..16): general Z formula, Z >= 18 only (the paper states E-A
          is negligible below that and gives no formula there).
        + Ca (Z=20, NoElectrons=20 or 19) and Fe (Z=26, NoElectrons=23 or 22): 4 explicit "other species"
          fits (Appendix A.4).
    Any other (Z, NoElectrons) combination, or approx != Arnaud1985EA(), raises an informative error rather
    than silently returning zero. An alpha::Float64 [a.u.] is returned.
    Quantity: a rate coefficient [cm^3/s] -- multiply by the electron number density n_e [1/cm^3] for the
        rate per ion [1/s].
source
JenaAtomicCalculator.Empirical.exponentialIntegralF1 — Method

Empirical.exponentialIntegralF1(x::Float64) ... to evaluate f1(x) = e^x int_1^infty dt/t e^(-xt) [i.e. e^x E1(x), with E1 the standard exponential integral], the special function used throughout the semi-empirical ionization and excitation- autoionization rate formulas of M. Arnaud & R. Rothenflug, A&AS 60, 425 (1985), their Appendix B.1. That paper gives f1(x) as three piecewise rational/logarithmic approximations (claimed accurate to better than 1%); a hand-transcription of the middle (0.02 < x < 10) piece was tried here first but found to be wrong throughout that range (not merely at an isolated point) when cross-checked against the exact function – e.g. it returned negative values for x in (0.02, 1) and diverged at x = 1, where f1(x) is in fact smooth and positive (f1(1) = 0.596). Rather than keep guessing at the exact printed form from a scanned 1985 paper, f1(x) is evaluated here EXACTLY via SpecialFunctions.expint(x) (already a dependency of this module), i.e. f1(x) = exp(x) * SpecialFunctions.expint(x) – mathematically exact to floating-point precision, not an approximation, and eliminating this transcription risk entirely. A Float64 is returned.

source
JenaAtomicCalculator.Empirical.fluorescenceYield — Method

Empirical.fluorescenceYield(Z::Int64; data::PeriodicTable.AbstractYieldData=PeriodicTable.KrauseAdopted2016()) ... to provide the K-shell fluorescence yield omega_K of the neutral element with nuclear charge Z, i.e. the probability that a K-shell vacancy is filled by a radiative (K X-ray) rather than a radiationless (Auger) transition. A yield::Float64 in [0,1] is returned; cf. Empirical.augerYield for its complement.

source
JenaAtomicCalculator.Empirical.forbiddenExcitationCrossSection — Method

Empirical.forbiddenExcitationCrossSection(energies::Array{Float64,1}, iConf::Configuration, fConf::Configuration, approx::Empirical.ConstantCollisionStrength; printout::Bool=false, data::PeriodicTable.AbstractEnergyData=PeriodicTable.XrayDataBooklet()) ... to estimate empirically the electron-impact excitation (EIE) cross section for the optically forbidden (M1 or E2) transition iConf -> fConf at the given (incident free-electron) energies, by assuming a constant, energy-independent collision strength Omega = approx.Omega (of order unity by convention when no detailed calculation is available), cf. Empirical.ConstantCollisionStrength. The standard collision-strength cross section [e.g. D. Osterbrock & G. Ferland, Astrophysics of Gaseous Nebulae and Active Galactic Nuclei, 2nd ed., University Science Books (2006)] sigma^(EIE) (eps) = pi / (2 eps) * Omega / gi, with eps the incident electron energy and gi the statistical weight of iConf, cf. Basics.extractFromConfiguration(Basics.Multiplicity(), iConf). A css::Array{Float64,1} [a.u.] is returned. Quantity: a cross section [a.u.] – a property of the ion alone, independent of the plasma; fold it with the electron field, or pass the configurations to Empirical.forbiddenExcitationPlasmaAlpha, to obtain a rate.

    Note: the transition multipole is determined from Delta-l between the shells involved, exactly as in
          Empirical.photoemissionEinsteinA: Delta-l = 0 -> M1, Delta-l = 2 -> E2. An optically allowed (E1,
          Delta-l = 1) transition is rejected -- use Empirical.impactExcitationCrossSection(energies, iConf,
          fConf, VanRegemorter1962()) instead -- and Delta-l >= 3 (E3 and higher) raises an error, since no
          useful collision-strength estimate is attempted for those higher multipoles.
source
JenaAtomicCalculator.Empirical.forbiddenExcitationPlasmaAlpha — Method

Empirical.forbiddenExcitationPlasmaAlpha(eDist::Distribution.AbstractElectronDistribution, iConf::Configuration, fConf::Configuration; approx::Empirical.ConstantCollisionStrength=Empirical.ConstantCollisionStrength(), printout::Bool=false, data::PeriodicTable.AbstractEnergyData=PeriodicTable.XrayDataBooklet()) ... to estimate empirically the electron-impact excitation plasma rate coefficient alpha^(EIE) for the optically forbidden (M1 or E2) transition iConf -> fConf and an electron distribution eDist at the electron temperature Te, by folding Empirical.forbiddenExcitationCrossSection with the electron distribution, alpha^(EIE) (Te; i -> f) = int d(eps) fe(eps; Te) v(eps) sigma^(EIE)(eps). For a constant collision strength Omega and a Maxwellian eDist, this reduces analytically to the textbook formula alpha^(EIE) (Te; i -> f) = 8.629e-6 / (gi sqrt(Te[K])) * Omega * exp(-deltaE/Te) [cm^3/s] [e.g. D. Osterbrock & G. Ferland, Astrophysics of Gaseous Nebulae and Active Galactic Nuclei, 2nd ed., University Science Books (2006)], which served as an independent numerical cross check of the Gauss-Legendre folding used here. An alpha::Float64 [a.u.] is returned. Quantity: a rate coefficient [cm^3/s] – multiply by the electron number density n_e [1/cm^3] to obtain the rate per ion [1/s].

source
JenaAtomicCalculator.Empirical.freeFreeAbsorptionCoefficient — Method

Empirical.freeFreeAbsorptionCoefficient(omegas::Array{Float64,1}, Z::Int64, ne::Float64, kT::Float64) ... to provide the free-free (inverse-bremsstrahlung) absorption coefficient, PER ABSORBING ION of charge Z, at every photon energy in omegas [a.u.], for a Maxwellian electron gas of density ne [a.u.] at temperature kT [Hartree]:

        kappa_ff(omega) = ne * Z^2 * (8*pi)/(3*sqrt(3)) * alpha * omega^(-3) * (1-exp(-u)) * <gff(u,gamma2)>

    with u = omega/kT and gamma2 = Z^2*0.5/kT (Ry = 0.5 Hartree). Derived from first principles, every
    step checked directly against the typeset page of W. J. Karzas & R. Latter, ApJS 6, 167 (1961): their
    eq. (15) is the classical Kramers cross section, eq. (16)-(17) define the Gaunt factor g_ff as its
    quantum correction, and eq. (18)'s Boltzmann electron-energy distribution integrates, after the
    substitution x = E_i/kT, u = omega/kT, into EXACTLY their own thermally-averaged <gff> (their eq. 21,
    reproduced independently by Empirical.freeFreeGauntFactor's table) -- the clean cancellation of every
    kT factor in that substitution is itself a check that no step of the derivation lost a factor. The
    `(1-exp(-u))` stimulated-emission correction is the standard LTE factor, not from Karzas & Latter
    directly. `Z <= 0` (a neutral stage, no Coulomb field for a passing electron) returns an all-zero
    array rather than raising, since it is a legitimate stage charge in a mixture, just one with no
    free-free contribution. A kappaFF::Array{Float64,1} is returned, matching `omegas` in length.
source
JenaAtomicCalculator.Empirical.freeFreeGauntFactor — Method

Empirical.freeFreeGauntFactor(u::Float64, gamma2::Float64) ... to provide the thermally-averaged, frequency-resolved free-free (bremsstrahlung) Gaunt factor <gff(u,gamma^2)>, the quantum correction to the classical Kramers free-free cross section used by Empirical.freeFreeAbsorptionCoefficient. u = omega/kT is the scaled photon energy and gamma^2 = Z^2*Ry/kT the scaled ionic binding scale [P. A. M. van Hoof et al., MNRAS 444, 420 (2014), eq. (21)]. Looked up by bilinear interpolation, in log10(u) and log10(gamma^2), on Sutherland's (1998, MNRAS 300, 321) table FREEFREE_GAUNT_TABLE (module-Empirical-inc-freefree-gaunt-data.jl); arguments outside the table's own log10 range [-4,4] x [-4,4] are CLAMPED to the nearest edge rather than extrapolated, since the table gives no information on the function's behaviour beyond it. A gff::Float64 is returned.

source
JenaAtomicCalculator.Empirical.groundConfiguration — Method

Empirical.groundConfiguration(Z::Int64) ... to build the ground configuration of the NEUTRAL atom with nuclear charge Z by filling the shells in Madelung (aufbau) order, i.e. by increasing n+l and, for equal n+l, by increasing n. This reproduces [Ar] 4s^2 for Ca, [Kr] 5s^2 for Sr and [Kr] 4d^10 5s^2 5p^6 for Xe, none of which follow from a strict (n,l) ordering. It is used only to recognise a ground configuration, never to construct one for a computation, and the half-filled-shell anomalies (Cr, Cu, Nb, Mo, Pd, Ag, ...) are therefore acceptable: for those the recognition simply fails and the caller falls back on its own estimate. A conf::Configuration is returned.

source
JenaAtomicCalculator.Empirical.impactExcitationCrossSection — Method

Empirical.impactExcitationCrossSection(energies::Array{Float64,1}, iConf::Configuration, fConf::Configuration, approx::Empirical.VanRegemorter1962; printout::Bool=false, aSource::Empirical.AbstractEmpiricalApproximation=Empirical.ScaledHydrogenic(), data::PeriodicTable.AbstractEnergyData=PeriodicTable.XrayDataBooklet()) ... to estimate empirically the electron-impact excitation (EIE) cross section for the optically allowed transition iConf -> fConf at the given (free-electron) energies, by using Van Regemorter's (1962) formula sigma^(EIE) (eps) = 8 pi / sqrt(3) * pi ao^2 * (ERyd^2 / (eps deltaE)) * f * gbar(x), where f is the absorption oscillator strength of the transition, obtained from the (empirical) Einstein-A value via f = (gf/gi) A / (2 alpha^3 omega^2), and gbar the effective Gaunt factor, cf. Empirical.effectiveGauntFactor(). The approx = VanRegemorter1962() argument selects this formula (as opposed to, e.g., the forbidden-transition treatment of a different approx type); the separate aSource keyword fixes how the Einstein-A value itself is estimated, with both ScaledHydrogenic() and GivenEinsteinA(..) supported. A css::Array{Float64,1} [a.u.] is returned. Quantity: a cross section [a.u.] – a property of the ion alone, independent of the plasma; fold it with the electron field, or pass the configurations to Empirical.impactExcitationPlasmaAlpha, to obtain a rate.

    Note: Van Regemorter's formula applies to optically allowed (E1) transitions only and is accurate to about
          a factor 2; it fails for forbidden transitions, for which no estimate is returned (an error is raised).
source
JenaAtomicCalculator.Empirical.impactExcitationPlasmaAlpha — Method

Empirical.impactExcitationPlasmaAlpha(eDist::Distribution.AbstractElectronDistribution, iConf::Configuration, fConf::Configuration; approx::Empirical.VanRegemorter1962=Empirical.VanRegemorter1962(), printout::Bool=false, aSource::Empirical.AbstractEmpiricalApproximation=Empirical.ScaledHydrogenic(), data::PeriodicTable.AbstractEnergyData=PeriodicTable.XrayDataBooklet()) ... to estimate empirically the electron-impact excitation plasma rate coefficient alpha^(EIE) for the optically allowed transition iConf -> fConf and an electron distribution eDist at the electron temperature Te, by folding Van Regemorter's cross section with the electron distribution, alpha^(EIE) (Te; i -> f) = int d(eps) fe(eps; Te) v(eps) sigma^(EIE)(eps). The approx = VanRegemorter1962() argument selects this formula; the separate aSource keyword fixes how the Einstein-A value itself is estimated, with both ScaledHydrogenic() and GivenEinsteinA(..) supported. An alpha::Float64 [a.u.] is returned. Quantity: a rate coefficient [cm^3/s] – multiply by the electron number density n_e [1/cm^3] to obtain the rate per ion [1/s].

source
JenaAtomicCalculator.Empirical.impactIonizationBindingEnergy — Method

Empirical.impactIonizationBindingEnergy(iConf::Configuration, iShell::Shell, pSource::Empirical.AbstractEmpiricalApproximation; data::PeriodicTable.AbstractEnergyData=PeriodicTable.XrayDataBooklet()) ... the binding energy P of iShell in iConf that enters the Lotz (1967) EII/TBR formula, from either tabulated data (pSource = ScaledHydrogenic()) or a JAC mean-field calculation (pSource = UsingJAC()). The mean-field orbital is generated with Basics.ALField(), not the DFS field used elsewhere in this module for PI/PR/PE: cf. the note in Basics.perform(::Empirical.Computation) on why an AL-Field eigenvalue, not a DFS one, is the binding energy in the Koopmans sense that this formula's 1/P^2 scaling needs. A P::Float64 [a.u.] is returned.

source
JenaAtomicCalculator.Empirical.impactIonizationCrossSection — Method

Empirical.impactIonizationCrossSection(energies::Array{Float64,1}, iConf::Configuration, fConf::Configuration, approx::Empirical.Lotz1967; printout::Bool=false, pSource::Empirical.AbstractEmpiricalApproximation=Empirical.ScaledHydrogenic(), data::PeriodicTable.AbstractEnergyData=PeriodicTable.XrayDataBooklet()) ... to estimate empirically the electron-impact ionization (EII) cross section for a transition iConf -> fConf at the given (free-electron) energies, by using the simplified formula of Lotz (1967, 1968) sigma^(EII) (eps) = A xi ln(eps/P) / (eps P), A = 4.5e-14 cm^2 eV^2, where P is the binding energy of the ionized shell and xi the number of its equivalent electrons. The pSource keyword fixes how P itself is obtained: ScaledHydrogenic() (the default) takes it from tabulated data, cf. Empirical.scaledBindingEnergy(); UsingJAC() takes it from a JAC mean-field calculation instead, cf. Empirical.impactIonizationBindingEnergy() – the same Lotz formula, a real rather than a tabulated input. Lotz's full formula carries subshell constants ai = 2.9 ... 4.5e-14 cm^2 eV^2 and two further parameters (bi, c_i) that lower the near-threshold cross section of neutral atoms; the simplified form with A = 4.5e-14 applies best to ions and overestimates light neutral atoms by ~40% (accuracy +40/-30%). A css::Array{Float64,1} [a.u.] is returned. Quantity: a cross section [a.u.] – a property of the ion alone, independent of the plasma; fold it with the electron field, or pass the configurations to Empirical.impactIonizationPlasmaAlpha, to obtain a rate.

source
JenaAtomicCalculator.Empirical.impactIonizationPlasmaAlpha — Method

Empirical.impactIonizationPlasmaAlpha(eDist::Distribution.AbstractElectronDistribution, iConf::Configuration, fConf::Configuration; approx::Empirical.AbstractEmpiricalApproximation=Lotz1967(), printout::Bool=false, pSource::Empirical.AbstractEmpiricalApproximation=Empirical.ScaledHydrogenic(), data::PeriodicTable.AbstractEnergyData=PeriodicTable.XrayDataBooklet()) ... to estimate empirically the electron-impact ionization plasma rate coefficient alpha^(EII) for a transition iConf -> fConf and an electron distribution eDist at the electron temperature Te, by folding the Lotz cross section with the electron distribution, alpha^(EII) (Te; i -> f) = int d(eps) fe(eps; Te) v(eps) sigma^(EII)(eps). The pSource keyword is passed through to Empirical.impactIonizationCrossSection and fixes how the binding energy P is obtained: ScaledHydrogenic() (tabulated) or UsingJAC() (a JAC mean-field calculation). An alpha::Float64 [a.u.] is returned. Quantity: a rate coefficient [cm^3/s] – multiply by the electron number density n_e [1/cm^3] to obtain the rate per ion [1/s].

source
JenaAtomicCalculator.Empirical.inelasticHCollisionRateMatrix — Method

Empirical.inelasticHCollisionRateMatrix(T::Float64, reaction::Empirical.InelasticHReaction, energies::Array{Pair{Configuration,Float64},1}; energyLabel::String="externally supplied", printout::Bool=false, zerosGL::Int64=64) ... convenience method that builds the Array{Empirical.InelasticHChannel,1} directly from reaction – reduced mass via Empirical.reducedMassH, molecular symmetry/statistical weight via Empirical.activeShell + Empirical.statisticalWeight (assuming a closed-shell, Lion=0/Sion=0.0, entrance ion) – and dispatches to Empirical.inelasticHCollisionRateMatrix(T, channels, pstatIonic, Z, mu; ...). energies must contain a total-energy entry [a.u.] (via Empirical.energyOf, e.g. [Configuration("[Xe]") => 0.0, Configuration("[Xe] 6s^1") => -0.368]) for reaction.iConfIon and for every reaction.fConfIon; this module never computes these itself (see the module note at the top of this file) – energyLabel is a short caller-supplied description (e.g. "single-configuration Dirac-Fock SCF" or "NIST") echoed in the printout so the origin of the numbers is never silently lost. A named tuple, as returned by the low-level method, is returned.

source
JenaAtomicCalculator.Empirical.inelasticHCollisionRateMatrix — Method

Empirical.inelasticHCollisionRateMatrix(T::Float64, channels::Array{Empirical.InelasticHChannel,1}, pstatIonic::Float64, Z::Float64, mu::Float64; printout::Bool=false, zerosGL::Int64=64) ... to compute the full rate-coefficient matrix Kif(T) for all neutralization, ion-pair formation, excitation and de-excitation processes among the given list of atomic/ionic channels and the ionic A^(Z+1)+ + H^- channel, following Steps 3-4 of Belyaev & Yakovleva, A&A 608, A33 (2017): for each channel j, the neutralization rate K(ionic -> j) = pstatIonic * Empirical.neutralizationReducedRate(T, Ej, ...), Eq. (1), and the inverse ion-pair-formation rate K(j -> ionic) from Empirical.detailedBalanceRate, Eq. (3); for each pair (i,k) of channels with Ei > Ek (i less deeply bound, i.e. the higher-lying state), the de-excitation rate K(i -> k) = pstati * Empirical.deExcitationReducedRate(T, Ei, Ek, ...), Eq. (2), and the inverse excitation rate K(k -> i) from detailed balance. A named tuple (channels=channels, ionicToChannel::Array{Float64,1}=, channelToIonic::Array{Float64,1}=, channelToChannel::Array{Float64,2}=) is returned, where channelToChannel[i,k] is the rate for channels[i] -> channels[k] (the diagonal is left at 0.0). Quantity: all rate coefficients [a.u.]; multiply by (Defaults.convertUnits("length: from atomic to cm",1.0)^3 / Defaults.convertUnits("time: from atomic to sec",1.0)) for the conventional [cm^3/s] used throughout the non-LTE literature (as in the printout below).

    Also available as Empirical.inelasticHCollisionRateMatrix(T, reaction::Empirical.InelasticHReaction, energies; ...)
    below, which builds channels automatically from Nuclear.Model/Configuration input instead of requiring the caller
    to hand-build the Empirical.InelasticHChannel list.
source
JenaAtomicCalculator.Empirical.ionicCovalentCoupling — Method

Empirical.ionicCovalentCoupling(Ej::Float64, EHminus::Float64, Rj::Float64) ... to compute the (diabatic) ionic-covalent coupling matrix element Hionic,j(Rj) by the semiempirical one-electron-transition formula of Olson, Smith & Bauer [Appl. Opt. 10, 1848 (1971)], Hionic,j = [sqrt(-Ej) + sqrt(-EH-)] / sqrt(2) * sqrt(Ej EH-) * Rj * exp( -0.86 Rj [sqrt(-Ej) + sqrt(-EH-)] / sqrt(2) ) , Eq. (12) of Belyaev & Yakovleva, A&A 608, A33 (2017). A value::Float64 [a.u.] is returned.

source
JenaAtomicCalculator.Empirical.ionizationThreshold — Method

Empirical.ionizationThreshold(Z::Float64, sh::Shell, conf::Configuration, data::PeriodicTable.AbstractEnergyData) ... returns the photoionization threshold of shell sh for the ion described by conf, i.e. the energy that has to be supplied to remove one electron from that shell; an energy::Float64 > 0. [a.u.] is returned.

    FOR THE OUTERMOST OCCUPIED SHELL OF AN ION this is E(q+1) - E(q), the (q+1)-th successive ionization potential of the element, and
    it is taken
    from PeriodicTable.ionizationPotentials_Nist2025. That is the quantity the threshold IS, and it is what an X-ray tabulation cannot
    give: Empirical.bindingEnergy is indexed by Z alone and returns the NEUTRAL value whatever the charge state, so for Sr+ it yields
    5.573 eV -- essentially the Sr I -> Sr II potential of 5.695 eV -- where the true edge is the Sr II -> Sr III potential of
    11.030 eV, low by about a factor of two. An edge in the wrong place opens the continuum at photon energies at which the ion cannot
    in fact be ionized at all.

    The shell is identified as the outermost OCCUPIED one -- largest n, then largest l -- and deliberately not through
    Basics.ValenceShells, which lists the OPEN shells only and would therefore return nothing for a closed-shell ion such as
    Sr^2+ = [Kr], leaving every noble-gas-like stage on the neutral threshold.

    Everything else falls through to Empirical.scaledBindingEnergy unchanged: inner shells, for which the X-ray value is the right
    starting point and the charge-state shift is a small correction to a large number; neutral atoms, where the tabulation is already
    the neutral value; and any case NIST does not cover. The first two are correct by construction and stay silent. The third is not:
    the table holds every charge state up to Z = 90 and nothing above, so an ION of Z >= 91 falls back on a NEUTRAL threshold, and that
    case now WARNS rather than degrading silently.
source
JenaAtomicCalculator.Empirical.landauZenerProbability — Method

Empirical.landauZenerProbability(Ej::Float64, EHminus::Float64, Z::Float64, v::Float64) ... to compute the single-pass Landau-Zener nonadiabatic transition probability at the ionic-covalent crossing Rj (Empirical.nonadiabaticRadius) for a radial velocity v [a.u.] there, pj = exp( -2 pi Hionic,j^2 (Z+1) / [(EH- - Ej)^2 v] ) , Eq. (11) of Belyaev & Yakovleva, A&A 608, A33 (2017) (the Eq. (10) definition with the diabatic potentials of Eqs. (6)-(7) already substituted). A value::Float64 (0 <= p_j <= 1) is returned.

source
JenaAtomicCalculator.Empirical.maxOrbitalAngularMomentum — Method

Empirical.maxOrbitalAngularMomentum(availableEnergy::Float64, R::Float64, mu::Float64) ... to compute the maximum total angular momentum quantum number J for which the radial velocity at the crossing radius R remains real for the given availableEnergy [a.u.] (the collision energy plus/minus any energy defect relevant for the crossing under consideration, cf. Eqs. (13), (18), (19)), i.e. the largest (non-negative) integer J with J(J+1) <= 2 mu R^2 availableEnergy. Returns -1 if even J=0 is classically closed (availableEnergy <= 0), signalling an empty J range. An n::Int64 is returned.

source
JenaAtomicCalculator.Empirical.meanCharge — Method

Empirical.meanCharge(Z::Int64, subshell::Subshell, conf::Configuration) ... to estimate the mean charge Z as seen by an electron in subshell for an ion with the given configuation. This charge is obtained by inverting the hydrogenic relation bE = Z^(eff)^2 / (2 n^2) for the binding energy bE > 0. of the given subshell, i.e. Z^(eff) = n sqrt(2 bE). A charge::Float64 is returned.

source
JenaAtomicCalculator.Empirical.meanExcitationEnergy — Method

Empirical.meanExcitationEnergy(Z::Int64) ... to estimate the mean excitation energy Ibar of the bound electrons of a neutral atom with nuclear charge Z, Ibar = 9.1 Z (1 + 1.9 Z^(-2/3)) [eV] , following Segre (1977) and Roy & Reed (1968), as quoted by Milne et al. (1999, Eq. 2). This estimate agrees with the ICRU tabulations to typically 5-10% for medium and heavy elements (Al: 159 eV vs. the tabulated 166 eV) but becomes rough for the lightest atoms (H: 26 eV vs. the tabulated 19.2 eV); since Ibar only enters logarithmically, this affects the stopping power at the few-percent level. An energy::Float64 [a.u.] is returned.

source
JenaAtomicCalculator.Empirical.molecularOrbitalFraction — Method

Empirical.molecularOrbitalFraction(Lj::Int64, Lion::Int64) ... to compute the fraction of channel j's own (2*Lj+1) orbital magnetic sublevels that fall on a molecular Lambda value also reachable by the ionic entrance term (Lion combined with a partner of L=0, i.e. Lambda_ion = 0,1,...,Lion). Since both Lj and Lion are combined with an L=0 partner (H or H^-), their own Lambda ranges are 0,1,...,Lj and 0,1,...,Lion respectively, so the shared range is 0,1,...,min(Lj,Lion) – always including Lambda=0. A value::Float64 in (0,1] is returned.

source
JenaAtomicCalculator.Empirical.molecularSpinFraction — Method

Empirical.molecularSpinFraction(Sj::Float64, Sion::Float64) ... to compute the fraction of channel j's own 2(2Sj+1) spin magnetic sublevels (from combining the atomic spin Sj with neutral hydrogen's S=1/2) that fall on the molecular spin Smol = Sion required by the ionic entrance term (combining the ionic term's own spin Sion with H^-'s S=0, so the ionic side has the single, unsplit value Smol = Sion). Returns 0.0 if Sion is not among {Sj+1/2, Sj-1/2} (no one-electron-transfer path conserves spin between the two terms). A value::Float64 in [0,1] is returned.

source
JenaAtomicCalculator.Empirical.neutralizationCrossSection — Method

Empirical.neutralizationCrossSection(E::Float64, Ef::Float64, EHminus::Float64, Z::Float64, mu::Float64; printout::Bool=false) ... to compute the (two-channel-approximation) cross section for the neutralization process A^(Z+1)+ + H^- -> A^Z+(f) + H at collision energy E [a.u.], by summing the Landau-Zener transition probability P^Nif(J,E) = 2 pf (1-pf) (Eq. 9) over total angular momentum J, sigma^Nif(E) = pi / (2 mu E) * sumJ (2J+1) P^Nif(J,E) , Eq. (5) of Belyaev & Yakovleva, A&A 608, A33 (2017), with the radial velocity at the crossing from Eq. (13). A sigma::Float64 [a.u.] is returned. Quantity: a cross section [a.u.] – fold with a Maxwellian velocity distribution, or use Empirical.neutralizationReducedRate, to obtain a (reduced) rate coefficient.

source
JenaAtomicCalculator.Empirical.neutralizationReducedRate — Method

Empirical.neutralizationReducedRate(T::Float64, Ef::Float64, EHminus::Float64, Z::Float64, mu::Float64; printout::Bool=false, zerosGL::Int64=64, checkRadius::Bool=true) ... to compute the reduced rate coefficient Nif(T; Ef) for the neutralization process A^(Z+1)+ + H^- -> A^Z+(f) + H at temperature T [a.u.], by folding Empirical.neutralizationCrossSection with the Maxwell distribution, Nif(T; Ef) = sqrt(8/(pi mu (kT)^3)) int0^infty sigma^Nif(E) E exp(-E/kT) dE , Eq. (4) of Belyaev & Yakovleva, A&A 608, A33 (2017), evaluated by Gauss-Legendre quadrature over [0, 30 kT] (converged to well below 1% for this smooth, exponentially-cut-off integrand). A value::Float64 [a.u.] is returned. Quantity: a (reduced) rate coefficient [cm^3/s] – multiply by the statistical probability pstat of the initial (ionic) molecular state, Eq. (1), to obtain the actual rate coefficient Kif(T); cf. Empirical.detailedBalanceRate for the inverse (ion-pair formation) process.

    Validated against Belyaev & Yakovleva (2017): N_if(T=6000 K, Ef=-4.29573 eV) for Ba2+ + H- -> Ba+(6d) + H
    gives 7.588e-8 cm^3/s here vs. the paper's quoted 7.59e-8 cm^3/s -- agreement to 3 significant figures.
source
JenaAtomicCalculator.Empirical.nonadiabaticRadius — Method

Empirical.nonadiabaticRadius(Ej::Float64, EHminus::Float64, Z::Float64) ... to compute the internuclear distance Rj at which the ionic A^(Z+1)+ + H^- (Coulomb) diabatic potential crosses the covalent A^Z+(j) + H (flat) diabatic potential, Rj = (Z+1) / (EH- - Ej) , Eq. (8) of Belyaev & Yakovleva, A&A 608, A33 (2017). Ej and EHminus are electronic bound energies [a.u., negative]. A value::Float64 [a.u.] is returned.

source
JenaAtomicCalculator.Empirical.outermostShell — Method

Empirical.outermostShell(conf::Configuration) ... to determine the outermost occupied shell of the given configuration, i.e. the occupied shell of largest n and, among those, of largest l. For the ground configuration of a neutral atom this is the valence shell, the one from which the first ionization potential removes an electron; it is 3p for Ar, 4s for Ca, 4s for Fe (not 3d), 5s for Sr and 6s for Ce. A shell::Shell is returned; an empty configuration raises an error.

source
JenaAtomicCalculator.Empirical.photodeexcitationPlasmaRatePerIon — Method

Empirical.photodeexcitationPlasmaRatePerIon(dist::Distribution.AbstractPhotonDistribution, iConf::Configuration, fConf::Configuration; approx::Empirical.AbstractEmpiricalApproximation=UsingJAC(), printout::Bool=false)

... to estimate the total photodeexcitation plasma rate R^(PD: total, per ion) (T; i -> f) for a transition from 
    iConf -> fConf by applying a given photon distribution dist. A rate::Float64 is returned.
    Quantity: a rate per ion [1/s] in the given field -- multiply by the ion number density n_ion [1/cm^3] to obtain
        the volumetric rate [1/(cm^3 s)].
source
JenaAtomicCalculator.Empirical.photoemissionEinsteinA — Method

Empirical.photoemissionEinsteinA(iConf::Configuration, fConf::Configuration, approx::Empirical.ScaledHydrogenic; printout::Bool=false, data::PeriodicTable.AbstractEnergyData=PeriodicTable.XrayDataBooklet()) ... to estimate empirically the Einstein-A value for a transition from iConf -> fConf. The transition energy is obtained as the difference of the binding energies of the two shells involved, cf. Empirical.scaledBindingEnergy(). A named triple (multipole::EmMultipole=, energy::Float64=, rate::Float64=) is returned. Quantity: a spontaneous rate [1/s] – an intrinsic property of the ion, independent of the plasma.

    Rates are estimated for E1 and M1 transitions only:
    + E1 (Delta-l = 1): from a hydrogenic scaling of the r-expectation values; good to a factor ~1.5
      (Ne K-alpha: 6.0e13 vs. the literature 8.8e13 1/s).
    + M1 (Delta-l = 0, always Delta-n != 0 here): such transitions are non-relativistically forbidden and
      proceed only via the relativistically induced M1 amplitude; the estimate scales the hydrogenic
      2s -> 1s anchor W(M1) = 2.496e-6 Z^10 1/s (Breit & Teller 1940; Johnson 1972) with the transition
      energy, (Delta-E/0.375 Hartree)^5, and is good to an order of magnitude only.
    + E2 and higher multipoles raise an error: the crude <r>-proxies misjudge their line strengths by
      orders of magnitude, and no silent zero is returned; use GivenEinsteinA(..) or the PhotoEmission
      module instead.
source
JenaAtomicCalculator.Empirical.photoemissionEinsteinA — Method

Empirical.photoemissionEinsteinA(iConf::Configuration, fConf::Configuration, approx::Empirical.UsingJAC; printout::Bool=false) ... to estimate empirically the Einstein-A value for a transition from iConf -> fConf by using simple JAC computations of transition and Einstein rates. A named triple (multipole::EmMultipole=, energy::Float64=, rate::Float64=) is returned. Quantity: a spontaneous rate [1/s] – an intrinsic property of the ion, independent of the plasma.

source
JenaAtomicCalculator.Empirical.photoemissionEinsteinA — Method

Empirical.photoemissionEinsteinA(approx::Empirical.GivenEinsteinA; printout::Bool=false) ... to simply return the Einstein-A value for the given transition. A named (multipole::EmMultipole=, energy::Float64=, rate::Float64=) is returned. Quantity: a spontaneous rate [1/s] – an intrinsic property of the ion, independent of the plasma.

source
JenaAtomicCalculator.Empirical.photoexcitationPlasmaRatePerIon — Method

Empirical.photoexcitationPlasmaRatePerIon(dist::Distribution.AbstractPhotonDistribution, iConf::Configuration, fConf::Configuration; approx::Empirical.AbstractEmpiricalApproximation=UsingJAC(), printout::Bool=false)

... to estimate the photoexcitation plasma rate per ion R^(PX: per ion) (T; i -> f) for a transition from 
    iConf -> fConf by applying a given photon distribution dist. A rate::Float64 is returned.
    Quantity: a rate per ion [1/s] in the given field -- multiply by the ion number density n_ion [1/cm^3] to obtain
        the volumetric rate [1/(cm^3 s)].
source
JenaAtomicCalculator.Empirical.photoionizationCrossSection — Method

Empirical.photoionizationCrossSection(omega::Float64, iConf::Configuration, fConf::Configuration; approx::Empirical.AbstractEmpiricalApproximation=ScaledHydrogenic(), printout::Bool=false)

... to estimate empirically the PI cross section for a transition from iConf -> fConf by applying some simple approximation 
    as determined by approx. For printout=true, basic information are printed about the input parameters, approximation 
    as well as the results in user-defined units. A cs::Float64 [in a.u.] is returned.
    Quantity: a cross section [a.u.] -- a property of the ion alone, independent of the plasma; fold it with the
        photon/electron field, or pass the configurations to a *PlasmaRatePerIon / *PlasmaAlpha function, to obtain a rate.
source
JenaAtomicCalculator.Empirical.photoionizationCrossSection — Method

Empirical.photoionizationCrossSection(omegas::Array{Float64,1}, iConf::Configuration, fConf::Configuration, approx::Empirical.ScaledHydrogenic; printout::Bool=false, data::PeriodicTable.AbstractEnergyData=PeriodicTable.XrayDataBooklet()) ... to estimate empirically the photoionization cross sections for a transition from iConf -> fConf by using the binding energy (ionization potential) of the ionized shell, cf. Empirical.scaledBindingEnergy(), and Kramer's (1923) empirical formula. A css::Array{Float64,1} [a.u.] is returned. Quantity: a cross section [a.u.] – a property of the ion alone, independent of the plasma; fold it with the photon/electron field, or pass the configurations to a *PlasmaRatePerIon / *PlasmaAlpha function, to obtain a rate.

source
JenaAtomicCalculator.Empirical.photoionizationCrossSection — Method

Empirical.photoionizationCrossSection(omegas::Array{Float64,1}, iConf::Configuration, fConf::Configuration, approx::Empirical.UsingJAC; printout::Bool=false) ... to estimate the photoionization cross section for a transition from iConf -> fConf from mean-field (DFS) orbitals and single-electron E1 amplitudes, sigma^(PI) (omega) = Nshell/(4l+2) * 8 pi^3/(alpha omega) * sum(subshells j = l+-1/2) sum(E1 channels kappac) |<eps kappac || O^(E1) || n kappa>|^2, with energy-normalized continuum orbitals in the local (nuclear + DFS) potential; the prefactor 8 pi^3/(alpha omega) belongs to the Johnson (2007) convention of the reduced emission matrix elements and is consistent with the (well-tested) bound-bound Einstein rate A = 8 pi alpha omega/(2Ji+1) |<f||O||i>|^2 by smearing a line into the continuum. A css::Array{Float64,1} [a.u.] is returned. Quantity: a cross section [a.u.] – a property of the ion alone, independent of the plasma; fold it with the photon/electron field, or pass the configurations to a *PlasmaRatePerIon / *PlasmaAlpha function, to obtain a rate.

    Note: Validated against experiment: He 1s^2 gives 6.3/4.9/1.7/0.32 Mb at 26/30/50/100 eV against the measured
          ~7.4/5.5/2/0.3 Mb (within ~15%); Ne 2p^6 reproduces shape and magnitude to ~30-50%. The *thresholds* are
          the mean-field (DFS) orbital energies and lie below the true ionization potentials (Ne 2p: 12.0 eV
          instead of 21.6 eV) -- a property of this approximation that has to be kept in mind near threshold.
          The continuum requires a radial grid with a linear tail; the exponential grid formerly used here
          produced spurious box resonances (92 Mb at 100 eV for Ne 2p) that invalidated all earlier results.
          The amplitudes further depend on the globally selected continuum method; the validation above refers
          to JAC's defaults, i.e. setDefaults("method: continuum, Galerkin") and
          setDefaults("method: normalization, Coulomb functions").
source
JenaAtomicCalculator.Empirical.photoionizationPlasmaRatePerIon — Method

Empirical.photoionizationPlasmaRatePerIon(dist::Distribution.AbstractPhotonDistribution, iConf::Configuration, fConf::Configuration; approx::Empirical.AbstractEmpiricalApproximation=ScaledHydrogenic(), printout::Bool=false, data::PeriodicTable.AbstractEnergyData=PeriodicTable.XrayDataBooklet()) ... to estimate empirically the photoionization plasma rate per ion R^(PI: per ion) (T; i -> f) for a transition from iConf -> fConf and a photon field dist at the radiation temperature T. A rate::Float64 [a.u.] is returned. Quantity: a rate per ion [1/s] in the given field – multiply by the ion number density n_ion [1/cm^3] to obtain the volumetric rate [1/(cm^3 s)].

    Note: In contrast to the photorecombination coefficient alpha^(PR) [cm^3/s], this quantity is a *rate* [1/s] and
          not a rate coefficient. The convolution R^(PI: per ion) = int d(omega) n(omega; T) c sigma^(PI)(omega)
          already contains the photon number density of the radiation field and, hence, needs no further
          multiplication by a density; it is the number of photoionization events per ion and per unit time.
          The free-electron density, in contrast, does not enter the PI process at all.
source
JenaAtomicCalculator.Empirical.photorecombinationCrossSection — Method

Empirical.photorecombinationCrossSection(energy::Float64, iConf::Configuration, fConf::Configuration; approx::Empirical.AbstractEmpiricalApproximation=ScaledHydrogenic(), printout::Bool=false)

... to estimate empirically the PR cross section for a transition from iConf -> fConf by applying some simple approximation 
    as determined by approx. For printout=true, basic information are printed about the input parameters, approximation 
    as well as the results in user-defined units. A cs::Float64 [in a.u.] is returned.
    Quantity: a cross section [a.u.] -- a property of the ion alone, independent of the plasma; fold it with the
        photon/electron field, or pass the configurations to a *PlasmaRatePerIon / *PlasmaAlpha function, to obtain a rate.
source
JenaAtomicCalculator.Empirical.photorecombinationCrossSection — Method

Empirical.photorecombinationCrossSection(energies::Array{Float64,1}, iConf::Configuration, fConf::Configuration, approx::Empirical.ScaledHydrogenic; printout::Bool=false, data::PeriodicTable.AbstractEnergyData=PeriodicTable.XrayDataBooklet()) ... to estimate empirically the (spontaneous) photorecombination cross section for a transition from iConf -> fConf by using the Einstein-Milne relation and the binding energy (ionization potential) of the captured electron, cf. Empirical.scaledBindingEnergy(), together with Kramer's (1923) empirical formula. A css::Array{Float64,1} [a.u.] is returned. Quantity: a cross section [a.u.] – a property of the ion alone, independent of the plasma; fold it with the photon/electron field, or pass the configurations to a *PlasmaRatePerIon / *PlasmaAlpha function, to obtain a rate.

source
JenaAtomicCalculator.Empirical.photorecombinationCrossSection — Method

Empirical.photorecombinationCrossSection(energies::Array{Float64,1}, iConf::Configuration, fConf::Configuration, approx::Empirical.UsingJAC; printout::Bool=false) ... to estimate the (spontaneous) photorecombination cross section for a transition from iConf -> fConf by applying the Einstein-Milne relation to the mean-field (UsingJAC) PI cross sections of the inverse process; the binding energy of the captured electron is taken from the mean-field orbital of fConf and is, therefore, consistent with the thresholds of those PI cross sections. A css::Array{Float64,1} [a.u.] is returned. Quantity: a cross section [a.u.] – a property of the ion alone, independent of the plasma; fold it with the photon/electron field, or pass the configurations to a *PlasmaRatePerIon / *PlasmaAlpha function, to obtain a rate.

    Note: cf. the validation notes of Empirical.photoionizationCrossSection(..., UsingJAC); in particular, the
          mean-field thresholds lie below the true ionization potentials.
source
JenaAtomicCalculator.Empirical.photorecombinationPlasmaAlpha — Method

Empirical.photorecombinationPlasmaAlpha(dist::Distribution.AbstractElectronDistribution, iConf::Configuration, fConf::Configuration; approx::Empirical.AbstractEmpiricalApproximation=ScaledHydrogenic(), printout::Bool=false) ... to estimate empirically the spontaneous PR plasma rate coefficient alpha^(PR: spontaneous) for a transition from iConf -> fConf by applying some simple approximation as determined by approx. For printout=true, basic information are printed about the approximation as well as the results. An alpha::Float64 is returned. This is a short-cut to the method above for a vanishing (vacuum) photon field. Quantity: a rate coefficient [cm^3/s] – multiply by the electron number density n_e [1/cm^3] to obtain the rate per ion [1/s].

source
JenaAtomicCalculator.Empirical.photorecombinationPlasmaAlpha — Method

Empirical.photorecombinationPlasmaAlpha(eDist::Distribution.AbstractElectronDistribution, pDist::Distribution.AbstractPhotonDistribution, iConf::Configuration, fConf::Configuration; approx::Empirical.AbstractEmpiricalApproximation=ScaledHydrogenic(), printout::Bool=false, data::PeriodicTable.AbstractEnergyData=PeriodicTable.XrayDataBooklet()) ... to estimate empirically the total PR plasma rate coefficient alpha^(PR: total), i.e. the sum of the spontaneous and the stimulated contribution, for a transition from iConf -> fConf. The free electrons follow the distribution eDist at the electron temperature Te, while the ambient radiation field is described by pDist at the (independent) radiation temperature T. An alpha::Float64 [a.u.] is returned. Quantity: a rate coefficient [cm^3/s] – multiply by the electron number density ne [1/cm^3] to obtain the rate per ion [1/s].

    Note: The stimulated enhancement [1 + nbar(omega)] is evaluated *inside* the integral over the electron energies.
          Photorecombination is a bound-free process, for which the emitted photon energy omega = eps + bEnergy is not
          a single number but varies across the whole electron distribution; there is, therefore, no unique omega at
          which a common factor [1 + nbar] could be taken out of the integral. Such a factorization is exact only for
          bound-bound processes (PD, PX), where omega = E_f - E_i is fixed. It would introduce an error that grows with
          the departure of the radiation field from the electron temperature, i.e. just in the non-LTE regime.
          For pDist = PhotonVacuumField, nbar = 0 and the spontaneous coefficient is recovered.
source
JenaAtomicCalculator.Empirical.recombinationConfigurations — Method

Empirical.recombinationConfigurations(conf::Configuration; nLayers::Int64=10) ... to enumerate the configurations that arise from conf by the capture of one additional electron. These are the capture channels of (three-body or radiative) recombination: every shell of conf with a vacancy as well as all empty (Rydberg) shells up to nmax = nvalence + nLayers, where n_valence is the largest principal quantum number occupied in conf. The enumeration truncates there – the contribution of a capture channel falls off with n, but the infinite Rydberg sum would diverge logarithmically, and very high Rydberg states are pressure-ionized in any real plasma. Only shells with l <= 3 (s, p, d, f) are included, since Slater's screening rules, which fix the binding energies of untabulated shells, are defined up to f electrons. A confs::Array{Configuration,1} is returned.

source
JenaAtomicCalculator.Empirical.reducedMassH — Method

Empirical.reducedMassH(nm::Nuclear.Model; ionMass::Union{Nothing,Float64}=nothing) ... to compute the reduced mass of the projectile ion (nuclear charge nm.Z) with a hydrogen atom, in electron-mass atomic units. The projectile's mass is taken from PeriodicTable.getData("mass", nm.Z) (the standard atomic weight, in amu) by default; note that nm.mass itself is NOT usable here – it is a generic A ~ 2Z+0.005Z^2 formula used only to derive the nuclear radius, not a real isotope mass. Pass ionMass [amu] explicitly to use a specific isotope instead. A value::Float64 [a.u.] is returned. Note: this rate model is thermally averaged over energy and summed over many partial waves, which washes out most of the mass-dependence of any single collision; even a +/-20% mass error changes typical rate coefficients by well under 1% (checked explicitly for Ba+ + H). Getting the mass exactly right is far less important than getting the level energies right for this model.

source
JenaAtomicCalculator.Empirical.scaledBindingEnergy — Method

Empirical.scaledBindingEnergy(Z::Float64, sh::Shell, conf::Configuration, data::PeriodicTable.AbstractEnergyData) ... to provide the binding energy of an electron in shell sh of the given configuration conf; this energy scales all hydrogenic (Kramers) estimates below and, hence, fixes the photoionization threshold as well as the transition energies. The semi-empirical binding energies of the neutral atom are taken from the tabulation data whenever this data set covers Z and sh and conf is not itself a genuine few-electron ion (see below). Failing that, and if sh is the OUTERMOST occupied shell of a NEUTRAL atom, the tabulated first ionization potential is used, since it is that shell's binding energy by definition. Only if neither applies is a Slater-screened hydrogenic estimate bE = Z^(eff)^2 / (2 n^2) used and a warning issued; Z^(eff) then follows from Slater's (1930) rules for the remaining electrons of conf. An energy::Float64 > 0. [a.u.] is returned.

    Note: The middle step matters because the X-ray compilations are CORE-level tables and carry no valence shell at
          all. Without it the valence threshold of every atom whose outermost shell is an s shell fell through to the
          Slater estimate, which is poor there and errs in BOTH directions: Sr 5s came out at 4.421 eV against the
          true 5.695 eV (-22%), Na 3s at 7.317 against 5.139 (+42%), Ca 4s at 6.907 against 6.113 (+13%). All three
          are now exact. Where a shell-resolved tabulated value does exist it is still preferred, and those cases
          were already sub-percent (Ar 3p -0.4%, Kr 4p +0.7%, Xe 5p -0.2%), so nothing that worked has moved.

    Note: A pure hydrogenic estimate is quite unreliable for valence shells; for neutral Ne, for instance, it places
          the 2p threshold near 85 eV, while the tabulated (true) binding energy is 21.6 eV. The tabulated values
          should therefore be preferred whenever they are available.

    Note: The tabulation is indexed by Z alone and, hence, always returns the *neutral*-atom value regardless of
          conf.NoElectrons. This is correct for the common "spectator-omitted shorthand" usage, where conf lists
          only the shells taking part in a transition of an otherwise near-neutral atom (e.g. a K-alpha analog
          Configuration("1s^1 2p^6") for Ne, with the 2s^2 spectator shell left out); it silently fails, however,
          for a conf that literally describes a stripped few-electron ion, such as He+'s Configuration("1s^1").
          A neutral atom with only 1 or 2 electrons is simply H or He, for which the tabulation already gives the
          correct (neutral) value; conf.NoElectrons < Z with conf.NoElectrons <= 2 therefore unambiguously signals
          a genuine few-electron ion rather than a spectator-omitted shorthand, and is routed directly to the
          Slater-screened hydrogenic estimate. This estimate is exact for H-like ions (no other electron to
          screen); for He-like ions, Slater's constant same-group screening (sigma = 0.30) systematically
          overestimates the binding energy, from about +31% at Z=3 (Li+) down to a few percent by Z ~ 20, as
          the ion becomes increasingly hydrogenic and the fixed screening constant matters proportionally
          less (checked against NIST/CRC data for the He isoelectronic sequence, Z = 3...18).
source
JenaAtomicCalculator.Empirical.statisticalWeight — Method

Empirical.statisticalWeight(Lj::Int64, Sj::Float64, Lion::Int64, Sion::Float64) ... to compute the statistical probability pstatj (Belyaev & Yakovleva 2017, Eqs. 1-2) that channel j = A^Z+(j) + H correlates, by molecular-term symmetry, to the ionic entrance state A^(Z+1)+ + H^- of term (Lion,Sion), as the product of Empirical.molecularOrbitalFraction and Empirical.molecularSpinFraction. Returns 0.0 (no correlation, channel j is not accessible from this ionic entrance) if the spin fraction is zero; see the section note above for the derivation and its verification against Table 1 of the paper. A value::Float64 in [0,1) is returned.

    SCOPE. The counting is carried out in L and S, so it presumes LS coupling on both sides. That is genuinely valid for
    the one-valence-electron systems the model was formulated on -- Ba+ in the paper -- but NOT for an open-shell or
    heavy ion in intermediate coupling, where L and S are not good quantum numbers while J and parity still are. There
    the weights returned here are approximate in a way this function cannot detect or report. A J-primary formulation,
    combining J_j with H's fixed J = 1/2 by the same Omega-projection counting, would remove the assumption; it should
    reproduce these numbers wherever LS coupling is in fact valid.
source
JenaAtomicCalculator.Empirical.stoppingPower — Method

Empirical.stoppingPower(energies::Array{Float64,1}, projectile::Empirical.AbstractStoppingProjectile, material::Empirical.AbstractStoppingMaterial, approx::Empirical.Axelrod1980; printout::Bool=false) ... to estimate the stopping power - dE/dx of an electron or positron at the given kinetic energies [a.u.] in the given material, by using Axelrod's (1980) relativistic (Bethe-type) energy loss for each electron population t, - dE/dx = sumt nt (4 pi e^4/(me c^2 beta^2)) [ ln( sqrt(gamma-1) gamma beta me c^2 / (hbar omegat) ) + 1/2 ln 2 - (beta^2/12) (23/2 + 7/(gamma+1) + 5/(gamma+1)^2 + 2/(gamma+1)^3) ], with the Lorentz factor gamma = 1 + E/(me c^2) and hbar omegat the characteristic energy of the population: the plasma energy hbar omegap for free electrons ("plasma losses") and the mean excitation energy Ibar for bound electrons ("ionization and excitation losses"); cf. Milne et al. (1999, Eqs. 1-3). The bracket is clamped at zero. A sps::Array{Float64,1} [a.u.] is returned; an ion projectile raises an error. Quantity: a stopping power [Hartree/a_o] – the energy loss per unit path length; it already contains the electron densities of the material.

    Note: In the nonrelativistic limit the bracket reduces exactly to the quantal electron logarithm
          ln(2 E/(hbar omega_t)) of Bethe1931, since sqrt(gamma-1) gamma beta m_e c^2 -> sqrt(2) E and the
          1/2 ln 2 completes the factor 2; the relativistic terms matter above ~50-100 keV. The (beta^2/12)
          term is the positron-type correction as transcribed by Milne et al., who apply it to electrons and
          positrons alike ("ignoring small differences in the relativistic corrections"); no density-effect
          correction is included, so the values grow logarithmically at highly relativistic energies.
source
JenaAtomicCalculator.Empirical.stoppingPower — Method

Empirical.stoppingPower(energies::Array{Float64,1}, projectile::Empirical.AbstractStoppingProjectile, material::Empirical.AbstractStoppingMaterial, approx::Empirical.Bethe1931; printout::Bool=false) ... to estimate the stopping power - dE/dx of the given projectile at the given kinetic energies [a.u.] in the given material, by using the quantal (Bethe-type) Coulomb logarithm for each electron population t, - dE/dx = sumt nt (4 pi z^2 e^4/(me v^2)) ln( kappa me v^2 / (hbar omegat) ) , where v = sqrt(2E/M) is the projectile velocity and hbar omegat the characteristic energy of the population (hbar omegap for free, Ibar for bound electrons). For a heavy projectile kappa = 2 (Bethe's original logarithm); for an electron or positron projectile kappa = 1, since the maximum energy transfer between identical particles is E/2, which reduces the logarithm by ln 2 [ln(me v^2/(hbar omegat)) = ln(2 E/(hbar omegat)), cf. Kozma & Fransson 1992, Eq. 1]. Each logarithm is clamped at zero. A sps::Array{Float64,1} [a.u.] is returned. Quantity: a stopping power [Hartree/a_o] – the energy loss per unit path length; it already contains the electron densities of the material.

    Note: The quantal logarithm takes b_min from the de Broglie wavelength and applies for projectile velocities
          v > |z| e^2/hbar (in a.u.: v > |z|); for bound electrons additionally v must well exceed the orbital
          velocities, i.e. E >> Ibar. These are bare Born values: no shell, Barkas (z^3), Bloch or density-effect
          corrections and no relativistic terms (use Axelrod1980 for electrons/positrons above ~50 keV).
          Typical accuracy 10-30% well above the characteristic energy.
source
JenaAtomicCalculator.Empirical.stoppingPower — Method

Empirical.stoppingPower(energies::Array{Float64,1}, projectile::Empirical.AbstractStoppingProjectile, material::Empirical.AbstractStoppingMaterial, approx::Empirical.Bohr1913; printout::Bool=false) ... to estimate the stopping power - dE/dx of the given projectile at the given kinetic energies [a.u.] in the given material, by using Bohr's (1913) classical Coulomb logarithm for each electron population t, - dE/dx = sumt nt (4 pi z^2 e^4/(me v^2)) ln( 1.123 me v^3 / (|z| e^2 omegat) ) , where v = sqrt(2E/M) is the projectile velocity and hbar omegat the characteristic energy of the population (hbar omegap for free, Ibar for bound electrons). The constant 1.123 = 2 exp(-gammaE) carries Euler's constant gammaE = 0.5772. Each logarithm is clamped at zero. A sps::Array{Float64,1} [a.u.] is returned. Quantity: a stopping power [Hartree/ao] – the energy loss per unit path length; it already contains the electron densities of the material.

    Note: Bohr's logarithm is the *classical* limit, ln(b_max/b_min) with the adiabatic distance b_max = v/omega_t
          and the classical collision diameter b_min = |z| e^2/(m_e v^2); it applies when the latter exceeds the
          de Broglie wavelength, i.e. for projectile velocities v < |z| e^2/hbar (in a.u.: v < |z|), but still well
          above the thermal/orbital velocities of the target electrons. It refers to a heavy (distinguishable)
          projectile: for electrons the indistinguishability reduces the logarithm by ln 2, cf. KozmaFranson1992.
          Accuracy: a factor ~2 at best; no shell, Barkas (z^3), Bloch or density-effect corrections.
source
JenaAtomicCalculator.Empirical.stoppingPower — Method

Empirical.stoppingPower(energies::Array{Float64,1}, projectile::Empirical.AbstractStoppingProjectile, material::Empirical.AbstractStoppingMaterial, approx::Empirical.KozmaFranson1992; printout::Bool=false) ... to estimate the stopping power - dE/dx of an electron at the given kinetic energies [a.u.] in a free-electron gas, by using the piecewise electron loss function of Kozma & Fransson (1992, Eqs. 1-3), - dE/dx = ne (2 pi e^4/E) ln( 2 E / (hbar omegap) ) for E > 14 eV, - dE/dx = ne (2 pi e^4/E) ln( me v^3 / (gamma e^2 omegap) ) for kT << E < 14 eV, with gamma = exp(gammaE) = 1.7811 (gammaE = 0.5772, Euler's constant; Schunk & Hays 1971). The boundary at 14 eV is physical: it marks E = m e^4/(2 hbar^2) = 1 Ry, where the de Broglie wavelength overtakes the classical distance of closest approach. Both branches carry the same -ln 2 electron-projectile correction relative to their heavy-projectile (Bohr/Bethe) counterparts, since 1.123 x 1.7811 = 2. The logarithm is clamped at zero. A sps::Array{Float64,1} [a.u.] is returned; any other projectile or material raises an informative error, since this loss function is defined just for this one case. Quantity: a stopping power [Hartree/ao] – the energy loss per unit path length; it already contains the electron density n_e.

source
JenaAtomicCalculator.Empirical.stoppingPowerPrintout — Method

Empirical.stoppingPowerPrintout(projectile::Empirical.AbstractStoppingProjectile, material::Empirical.AbstractStoppingMaterial, sa::String, energies::Array{Float64,1}, targets::Array{Tuple{Float64,Float64,String},1}, sps::Array{Float64,1}) ... to print a common report about the given stopping powers: the projectile and material in words, the approximation and its regime of validity (sa), the electron populations that take up the energy, the neglected mechanisms, the energies in user-defined units, the stopping powers in eV/cm as well as – for atom gases with known mass number – the mass stopping power in MeV cm^2/g. Nothing is returned.

source
JenaAtomicCalculator.Empirical.stoppingProjectileData — Method

Empirical.stoppingProjectileData(projectile::Empirical.AbstractStoppingProjectile) ... to provide the parameters of a projectile as needed in the stopping-power formulas: its squared charge z^2, its mass M [m_e] and whether it is a lepton (electron/positron), for which the maximum energy transfer to a target electron is limited and the relativistic (Axelrod-type) loss applies. A tuple (z2::Float64, M::Float64, lepton::Bool) is returned.

source
JenaAtomicCalculator.Empirical.stoppingTargets — Method

Empirical.stoppingTargets(material::Empirical.AbstractStoppingMaterial) ... to decompose a material into its electron populations as needed in the stopping-power formulas; each population is characterized by its electron density and one characteristic energy: the plasma energy hbar omegap = sqrt(4 pi ne) [a.u.] for free electrons, and the mean excitation energy Ibar for bound electrons. A targets::Array{Tuple{Float64,Float64,String},1} of (density, energy, label) is returned.

source
JenaAtomicCalculator.Empirical.threeBodyRecombinationPlasmaAlpha — Method

Empirical.threeBodyRecombinationPlasmaAlpha(eDist::Distribution.ElectronMaxwell, iConf::Configuration, fConf::Configuration; approx::Empirical.AbstractEmpiricalApproximation=Lotz1967(), printout::Bool=false, pSource::Empirical.AbstractEmpiricalApproximation=Empirical.ScaledHydrogenic(), data::PeriodicTable.AbstractEnergyData=PeriodicTable.XrayDataBooklet()) ... to estimate empirically the three-body recombination (TBR) plasma rate coefficient alpha^(TBR) for the capture channel iConf + e + e -> fConf + e, by detailed balance with the (Lotz) electron-impact ionization of fConf. In LTE, the Saha equation fixes the ratio of the forward and backward rates, so that alpha^(TBR) (Te; i -> f) = alpha^(EII) (Te; f -> i) * gf/(2 gi) * (2 pi/Te)^(3/2) * exp(P/Te), with P the binding energy of the captured electron and the factor 2 the spin degeneracy of the free electron. The pSource keyword fixes how P itself is obtained: ScaledHydrogenic() (the default, tabulated) or UsingJAC() (a JAC mean-field calculation), cf. Empirical.impactIonizationBindingEnergy() – the same Saha/Lotz formula throughout, a real rather than a tabulated input. An alpha::Float64 [a.u.] is returned. Quantity: a three-body rate coefficient [cm^6/s] – multiply by the electron number density squared n_e^2 [1/cm^6] to obtain the rate per ion [1/s]; TBR therefore dominates over radiative recombination at high electron densities.

    Note: The two exponentials exp(+P/T_e) (Saha) and exp(-P/T_e) (threshold suppression of the EII coefficient)
          cancel analytically but *not* numerically: for T_e << P the EII coefficient underflows and the product
          becomes 0 * Inf. The Saha exponential is therefore folded into the integrand, where it combines to the
          well-behaved exp(-(eps - P)/T_e) on the threshold-started mesh. This cancellation requires a Maxwellian
          electron distribution -- detailed balance in this form holds only in (local) thermodynamic equilibrium --
          and the method is restricted accordingly.
source
JenaAtomicCalculator.Empirical.threeBodyRecombinationPlasmaAlpha — Method

Empirical.threeBodyRecombinationPlasmaAlpha(eDist::Distribution.ElectronMaxwell, iConf::Configuration; nLayers::Int64=10, approx::Empirical.AbstractEmpiricalApproximation=Lotz1967(), printout::Bool=false, pSource::Empirical.AbstractEmpiricalApproximation=Empirical.ScaledHydrogenic(), data::PeriodicTable.AbstractEnergyData=PeriodicTable.XrayDataBooklet()) ... to estimate empirically the total three-body recombination plasma rate coefficient alpha^(TBR) (Te; i) of the ion iConf, i.e. the sum over all capture channels iConf + e + e -> fConf + e as enumerated by Empirical.recombinationConfigurations(iConf; nLayers): every vacancy of iConf and all empty (Rydberg) shells with l <= 3 up to nmax = nvalence + nLayers, where the sum is truncated. The pSource keyword is passed through to each channel's Empirical.threeBodyRecombinationPlasmaAlpha(..., iConf, fConf, ...). An alpha::Float64 [a.u.] is returned. Quantity: a three-body rate coefficient [cm^6/s] – multiply by the electron number density squared ne^2 [1/cm^6] to obtain the rate per ion [1/s].

    Note: The truncation is not merely numerical but physical: capture into shells with binding energies well
          below T_e is undone by the next collision, so that including ever higher Rydberg shells would inflate
          the coefficient without describing effective recombination; n_max should be chosen in accordance with
          the plasma density (pressure ionization). As a consequence, this *truncated direct-capture* sum is not
          the classical collisional-radiative coefficient of Mansbach & Keck (1969): their scaling
          alpha^(TBR) ~ T_e^(-9/2) rests on the cascade through levels with binding energy ~ T_e, which lie above
          the truncation at low T_e. For H^+ with nLayers = 10, the present sum scales only ~ T_e^(-1.2 ... -1.4)
          between 0.05 and 0.4 eV and lies about an order of magnitude below the classical estimate at 0.1 eV.
source
JenaAtomicCalculator.Empirical.totalEnergy — Method

Empirical.totalEnergy(Z::Int64, conf::Configuration; data::PeriodicTable.AbstractEnergyData = PeriodicTable.XrayDataBooklet() ) ... to estimate the total energy of the given configuration for an atom or ions with nuclear charge Z; this total energy is determined either from the (negative) sum of all binding energies or the (negative) sum of ionization potentials that are needed to remove the remaining electrons from the configuration. A warning is issued if electrons occur in excited shells with regard to the ground configuration of the neutral atom. An energy::Float64 is returned.

    Note: This is an estimate of the total energy which need not to be so accurate.
source
JenaAtomicCalculator.Empirical.tunnelingIonizationRate — Method

Empirical.tunnelingIonizationRate(fields::Array{Float64,1}, iConf::Configuration, fConf::Configuration, approx::Empirical.ADK1986; m::Int64=0, printout::Bool=false, data::PeriodicTable.AbstractEnergyData=PeriodicTable.XrayDataBooklet()) ... to estimate the tunneling ionization rate of an electron from iConf -> fConf at the given (quasi-static) electric field strengths [a.u.]; a convenience wrapper around the direct (Z, Ip, l) method above, which determines the residual ion charge Z = Z_nuclear - fConf.NoElectrons and the binding energy Ip of the ionized shell via Empirical.scaledBindingEnergy(), then dispatches to the direct method. A rates::Array{Float64,1} [a.u.] is returned. Quantity: a field-ionization rate [1/s], cf. the direct (Z, Ip, l) method for the full description.

    Note: Empirical.scaledBindingEnergy() recognizes a genuine H-like or He-like ion (iConf.NoElectrons <= 2 and
          < Z_nuclear, e.g. He^+ = "1s^1") and then returns the exact (H-like) or approximate (He-like, Slater
          screening; overestimates the binding energy by ~5-30%, decreasing with Z) hydrogenic binding energy
          automatically, rather than the *neutral*-atom tabulated value. A residual gap remains for more highly
          charged ions with 3 or more electrons (e.g. a Li-like O^5+): such an iConf is indistinguishable, from
          the electron count alone, from the common "spectator-omitted shorthand" for a near-neutral atom (e.g.
          Configuration("1s^1 2p^6") for Ne), so the *neutral*-element tabulated value is still used and may
          differ substantially from the true one for a genuinely stripped few-electron ion in this range. For
          such ions, prefer the direct (Z, Ip, l) method with an explicitly supplied (e.g. hydrogenic or
          NIST-tabulated) Ip.
source
JenaAtomicCalculator.Empirical.tunnelingIonizationRate — Method

Empirical.tunnelingIonizationRate(fields::Array{Float64,1}, Z::Float64, Ip::Float64, l::Int64, approx::Empirical.ADK1986; m::Int64=0, printout::Bool=false) ... to estimate the tunneling ionization rate of an electron with binding energy Ip [a.u.] and orbital quantum number l, in the Coulomb tail of an ion of residual charge Z (the charge seen by the escaping electron at large distance, i.e. the charge of the daughter ion once this electron is gone), at the given (quasi-static) electric field strengths [a.u.]. This is the direct (Z, Ip, l) form of Ammosov, Delone & Krainov's (1986) formula, W(F) = C^2_(n* l) f(l,m) Ip (2 kappa^3/F)^(2n-|m|-1) exp( -2 kappa^3/(3F) ), with kappa = sqrt(2 Ip), the effective principal quantum number n* = Z/kappa, cf. Empirical.effectiveQuantumNumbers(). The effective orbital quantum number l* is set to zero throughout, following standard practice (e.g. Ammosov et al. themselves, and essentially all production strong-field codes; l* would otherwise require the quantum-defect structure of the atom, which is rarely available), while the real orbital quantum number l and the magnetic quantum number m (default 0, i.e. the field aligned along the symmetry axis of the initial orbital) enter the angular factor f(l,m), cf. Empirical.adkAngularFactor(). A rates::Array{Float64,1} [a.u.] is returned. Quantity: a field-ionization rate [1/s] – already evaluated at the given field; nothing further needs to be multiplied. It is not a spontaneous (zero-field) rate: the field dependence is nonperturbative (an essential singularity at F = 0) and cannot be linearized or summed with a zero-field rate.

    Note: This is the *static-field* (quasi-DC) rate. For a field oscillating as F(t) = F0 cos(omega t), the
          literature ADK rate is usually quoted after averaging over one optical cycle, which multiplies the
          static rate above by sqrt(3F/(pi kappa^3)) [Bauer & Mulser, Eq. (10)]; that cycle-averaged variant is
          not implemented here, and F should be understood as the (peak) field strength of interest, e.g. F0.

    Note: The formula is valid in the tunneling regime, F well below the critical (barrier-suppression) field
          F_c = kappa^3/16; above F_c the true ionization rate is known to be substantially overestimated (the
          potential barrier no longer exists at the classical level). No warning is raised automatically, since
          F/F_c is not itself pathological to compute, but it is reported whenever printout = true.

    Note: For Z = 1, Ip = 0.5 [a.u.] and l = 0 (the pure hydrogenic 1s case, where n* = 1 and l* = 0 by
          construction), this rate reduces *exactly* to the closed Landau tunneling formula (numerically verified
          to 8 digits) -- a strong, literature-independent cross check of the coefficient normalization used
          here, cf. Empirical.adkCoefficient().
source
JenaAtomicCalculator.Empirical.validateHydrogenConfiguration — Method

Empirical.validateHydrogenConfiguration(conf::Configuration) ... to check that conf is either H(1s^1) or H^-(1s^2); raises an informative error otherwise, since Empirical.InelasticHReaction is restricted to hydrogen collision partners by design (see the section note above).

source

Semi-empirical estimates

JenaAtomicCalculator.Semiempirical.estimate — Method

Semiempirical.estimate(::EstimateBindingEnergyLarkins1977, Z::Int64, conf::Configuration) ... provides an approximate total binding energy of a configuration from Larkins (1977); no relaxation effects between hole states are included; an energy::Float64 in Hartree is returned.

source
JenaAtomicCalculator.Semiempirical.estimate — Method

Semiempirical.estimate(::EstimateBindingEnergyLarkins1977, Z::Int64, sh::Subshell) ... provides the binding energy of a subshell electron from the Larkins (1977) tabulation; an energy::Float64 in Hartree is returned.

source
JenaAtomicCalculator.Semiempirical.estimate — Method

Semiempirical.estimate(::EstimateBindingEnergyNist2025, Z::Int64, q::Int64) ... returns the q-th successive ionization potential for element Z from the NIST (2025) database, https://physics.nist.gov/PhysRefData/ASD/ionEnergy.html; q = 1 gives the first ionization potential (outermost/valence electron of the neutral atom), q = 2 the second, and so on. Covers Z = 1..90; an energy::Float64 in Hartree is returned.

source
JenaAtomicCalculator.Semiempirical.estimate — Method

Semiempirical.estimate(::EstimateBindingEnergyWilliams2000, Z::Int64, sh::Subshell) ... provides the binding energy of a subshell electron from the Williams et al. (2000) tabulation, https://userweb.jlab.org/~gwyn/ebindene.html; an energy::Float64 in Hartree is returned.

source
JenaAtomicCalculator.Semiempirical.estimate — Method

Semiempirical.estimate(::EstimateBindingEnergyXrayDataBooklet, Z::Int64, conf::Configuration) ... provides an approximate total binding energy of a configuration from the X-ray Data Booklet tabulation; no relaxation effects between hole states are included; an energy::Float64 in Hartree is returned.

source
JenaAtomicCalculator.Semiempirical.estimate — Method

Semiempirical.estimate(::EstimateBindingEnergyXrayDataBooklet, Z::Int64, sh::Subshell) ... provides the binding energy of a subshell electron from the X-ray Data Booklet tabulation; covers 1s1/2 through 6p3/2 (24 subshells); an energy::Float64 in Hartree is returned.

source
JenaAtomicCalculator.Semiempirical.estimate — Method

Semiempirical.estimate(::EstimateIonizationPotentialInnerShell, sh::Subshell, Z::Int64) ... estimates the ionization potential = binding energy of an electron in the given subshell; values are taken from Williams et al. (2000); covers 1s1/2 through 4p3/2 (12 subshells); an energy value::Float64 in Hartree is returned.

source
JenaAtomicCalculator.Semiempirical.estimateBindingEnergies — Method

Semiempirical.estimateBindingEnergies(Z::Float64, coreConf::Configuration, nRange::UnitRange{Int64}, l::Int64) ... to estimate the binding energies of the high-n shells with n = nRange and orbital angular momentum l for a (multiply-charged) ion with nuclear charge Z and core configuration coreConfiguration; Slater's (1930) screening rules are used to compute Z_eff for each (n, l) shell. A list of energies::Array{Float64} [in Hartree] is returned.

source
JenaAtomicCalculator.Semiempirical.estimateBindingEnergies — Method

Semiempirical.estimateBindingEnergies(Z::Float64, coreConf::Configuration, nRange::UnitRange{Int64}) ... to estimate the binding energies of the high-n shells with n = nRange for a (multiply-charged) ion with nuclear charge Z and core configuration coreConfiguration. A list of energies::Array{Float64} [in Hartree] is returned.

source
JenaAtomicCalculator.Semiempirical.estimateRadialExpectation — Method

Semiempirical.estimateRadialExpectation(Z::Float64, coreConf::Configuration, nRange::UnitRange{Int64}, l::Int64) ... to estimate the radial extent (r-expectation value) of the high-n shells with n = nRange and orbital angular momentum l for a (multiply-charged) ion with nuclear charge Z and core configuration coreConfiguration. A list of radii::Array{Float64} [in a_o] is returned.

source
JenaAtomicCalculator.Semiempirical.estimateSlaterZeff — Method

Semiempirical.estimateSlaterZeff(Z::Float64, conf::Configuration, sh::Subshell) ... estimates the effective nuclear charge Z_eff for a single electron in subshell sh, given the remaining electrons in conf and the nuclear charge Z. Slater's (1930) empirical screening rules are applied: same-group electrons contribute σ = 0.35 (0.30 for [1s]), electrons with n' = n−1 contribute σ = 0.85 (sp targets only), and all inner electrons contribute σ = 1.00. For d and f electrons all inner electrons contribute σ = 1.00. The configuration conf must represent the other N−1 electrons only (not include the target electron itself); a Zeff::Float64 is returned.

source
JenaAtomicCalculator.Semiempirical.noEntry — Method

Semiempirical.noEntry(source::String, sh::Shell, Z::Int64) ... raises because the named compilation of measured binding energies has no entry for this shell of this element; it never returns. The three tables mark a missing entry with -1., and all three estimate(::EstimateBindingEnergy..., ...) methods meet it identically, so they share ONE message rather than carrying twenty-eight error("stop aa") markers between them, as they did until 29-Aug-2026. A missing entry is an ordinary user request, not an internal fault: these compilations are incomplete for many elements, so the message says which table, which shell and which Z, and names the alternatives.

source