Eve library

phase-124.md

Tracker

Revision 1

On this page

Historical task source. The Eve task board holds current task state.

<!-- prettier-ignore-start -->

Phase 124: Kalika Spectroscopy Simulation — XAS/XANES/EXAFS, Photoemission, Vibrational Spectroscopy, NMR, Mossbauer & Electron Energy Loss#

Simulation of all major experimental spectroscopies from first principles, enabling direct comparison between computed and measured spectra. This phase covers X-ray absorption (XANES/EXAFS), photoemission (XPS/ARPES), vibrational (IR/Raman), nuclear magnetic resonance, Mossbauer, electron energy loss, and auxiliary spectroscopies (EPR, RIXS, XES, PL). Targets parity with FEFF (XAS), OCEAN (BSE-XAS), exciting (optics), Gaussian/ORCA (NMR), and WIEN2k spectroscopy modules. Tech stack: Rust (~35%) for compute-intensive kernels (multiple scattering, Raman tensors, EELS cross-sections), TypeScript (~55%) for orchestration/types/API, Python (~10%) for legacy format I/O and prototyping.


124.1 Scaffolding & Core Types (@kalika/spectroscopy)#

124.1.1 Library Scaffolding#

  • 124.1.1.1 Create libs/kalika/spectroscopy/ directory with project.json, package.json, tsconfig.json, and tsconfig.lib.json. Add @kalika/spectroscopy path mapping to tsconfig.base.json
  • 124.1.1.2 Create barrel export src/index.ts exporting all public types, classes, and factory functions from the library
  • 124.1.1.3 Set up Rust workspace member libs/kalika/spectroscopy/rust/ with Cargo.toml, configure napi-rs bindings for TypeScript interop, and add wasm-pack target for browser builds
  • 124.1.1.4 Create vitest configuration for the library with test fixtures directory __fixtures__/ containing sample spectra in common formats (XDI, VAMAS, JCAMP-DX, CSV)

124.1.2 Core Spectrum Types#

  • 124.1.2.1 Define Spectrum base interface: energy: Float64Array, intensity: Float64Array, metadata: SpectrumMetadata (source, date, units, broadening applied, edge/nucleus info), peaks: Peak[], clone(): Spectrum, slice(eMin, eMax): Spectrum
  • 124.1.2.2 Define Peak type: position: number, height: number, fwhm: number, area: number, shape: PeakShape, assignment: string, confidence: number
  • 124.1.2.3 Define PeakShape enum and associated parameter types: Gaussian, Lorentzian, Voigt, PseudoVoigt, DoniachSunjic, AsymmetricGaussian, each with shape-specific parameters (mixing ratio eta for pseudo-Voigt, asymmetry alpha for Doniach-Sunjic, etc.)
  • 124.1.2.4 Define XASSpectrum extending Spectrum: edge: XASEdge (K, L1, L2, L3, M1..M5), absorberElement: Element, absorberSite: number, mu: Float64Array (absorption coefficient), chi: Float64Array | null (EXAFS signal), chiK: Float64Array | null, chiR: Float64Array | null, e0: number, normalization: NormParams
  • 124.1.2.5 Define XPSSpectrum extending Spectrum: coreLevel: string (e.g., "C 1s", "O 1s"), bindingEnergy: Float64Array, kineticEnergy: Float64Array, photonEnergy: number, workFunction: number, shirleyBackground: Float64Array | null
  • 124.1.2.6 Define RamanSpectrum extending Spectrum: ramanShift: Float64Array (cm^-1), excitationWavelength: number (nm), temperature: number, polarization: RamanPolarization | null, stokesAntiStokes: 'stokes' | 'anti-stokes' | 'both'
  • 124.1.2.7 Define IRSpectrum extending Spectrum: wavenumber: Float64Array (cm^-1), transmittance: Float64Array | null, absorbance: Float64Array | null, mode: 'transmission' | 'atr' | 'reflectance'
  • 124.1.2.8 Define NMRSpectrum extending Spectrum: chemicalShift: Float64Array (ppm), nucleus: NMRNucleus, fieldStrength: number (Tesla), referenceCompound: string, lineShape: 'gaussian' | 'lorentzian' | 'powder' | 'mas', spinRate: number | null (Hz, for MAS)
  • 124.1.2.9 Define EELSSpectrum extending Spectrum: energyLoss: Float64Array (eV), region: 'low-loss' | 'core-loss', collectionAngle: number (mrad), convergenceAngle: number (mrad), beamEnergy: number (keV), zeroLossPeak: { position: number; fwhm: number } | null
  • 124.1.2.10 Define MossbauerSpectrum extending Spectrum: velocity: Float64Array (mm/s), transmission: Float64Array, nucleus: MossbauerNucleus, sourceIsotope: string, temperature: number (K), hyperfineParams: MossbauerHyperfineParams | null

124.1.3 Core Enums and Constants#

  • 124.1.3.1 Define XASEdge enum with all standard edges: K, L1, L2, L3, M1..M5, N1..N7, O1..O5. Include edgeEnergy(element: Element, edge: XASEdge): number lookup from tabulated X-ray edge energies (Henke/Elam database values)
  • 124.1.3.2 Define NMRNucleus type covering common nuclei: H1, C13, N15, O17, F19, Na23, Al27, Si29, P31, V51, Fe57, Cu63, Sn119, Pt195, etc. Include spin: number, gyromagneticRatio: number, naturalAbundance: number, quadrupoleMoment: number | null for each
  • 124.1.3.3 Define MossbauerNucleus enum: Fe57, Sn119, Eu151, Dy161, Au197, Np237, etc. Include excitedStateEnergy: number (keV), groundStateSpin: number, excitedStateSpin: number, naturalLinewidth: number (mm/s) for each isotope
  • 124.1.3.4 Define photoionization cross-section tables: implement photoionizationCrossSection(element: Element, orbital: string, photonEnergy: number): number using Yeh-Lindau or Scofield tabulated values for XPS and valence band cross-section weighting

124.1.4 Spectrum Container and Registry#

  • 124.1.4.1 Implement SpectrumCollection container: store multiple spectra with labels, group by type/edge/element, provide iteration, filtering by metadata, and merge operations (sum, average, weighted sum)
  • 124.1.4.2 Implement SpectrumRegistry singleton: register computed spectra by structure ID + spectroscopy type, cache results, provide lookup by structure hash, invalidate on structure change
  • 124.1.4.3 Implement spectrum serialization: toJSON() / fromJSON() for all spectrum types, with Float64Array encoded as base64 for compact storage. Include format version for forward compatibility

124.2 X-ray Absorption Spectroscopy — XAS (@kalika/xas)#

124.2.1 XAS Library Scaffolding#

  • 124.2.1.1 Create libs/kalika/xas/ directory with project.json, package.json, tsconfig.json. Add @kalika/xas path mapping to tsconfig.base.json
  • 124.2.1.2 Create Rust crate libs/kalika/xas/rust/ for performance-critical scattering computations: muffin-tin potentials, phase shifts, Green's function calculations. Configure napi-rs bindings

124.2.2 Muffin-Tin Potential Construction#

  • 124.2.2.1 Implement muffin-tin potential construction: given atomic positions and species, partition space into non-overlapping atomic spheres and interstitial region. Compute spherically-averaged potential inside each sphere from overlapping free-atom charge densities (Mattheiss prescription). Average potential in interstitial to define muffin-tin zero
  • 124.2.2.2 Implement self-consistent muffin-tin potential iteration: solve radial Schrodinger equation inside each sphere for the initial potential, compute new charge density from wavefunctions, mix old and new potentials (linear or Anderson mixing), iterate to self-consistency with convergence threshold on potential change (< 10^-4 Ry)
  • 124.2.2.3 Implement Hedin-Lundqvist self-energy correction: add complex self-energy to the muffin-tin potential to account for inelastic losses and finite photoelectron lifetime. Compute energy-dependent imaginary part from the electron gas model

124.2.3 Phase Shift Computation (Rust)#

  • 124.2.3.1 Implement radial Schrodinger equation solver (Rust): Numerov method for regular and irregular solutions inside each muffin-tin sphere at complex energy E + iGamma. Compute radial wavefunctions R_l(r) for angular momentum l = 0..l_max (l_max ~ 4 for K-edge, ~6 for L-edge)
  • 124.2.3.2 Implement scattering phase shift extraction: match radial wavefunction at sphere boundary to spherical Bessel/Neumann functions to extract complex phase shifts delta_l(E). Store as PhaseShiftTable { element: Element, lMax: number, energies: Float64Array, phases: Complex64Array[] }
  • 124.2.3.3 Implement scattering t-matrix from phase shifts: t_l = -sin(delta_l) exp(i*delta_l) / k, for use in multiple scattering equations. Handle complex phase shifts for absorptive potentials

124.2.4 XANES — Multiple Scattering Theory (Real-Space)#

  • 124.2.4.1 Implement free-electron propagator G0 between two sites: G0*{LL'}(R_i, R_j) = -ik * sum_L'' C{LL'L''} hl''(kR{ij}) Y_L''(R_ij_hat), where C are Gaunt coefficients, h_l are spherical Hankel functions, and L = (l,m) are angular momentum indices
  • 124.2.4.2 Implement full multiple scattering (FMS) matrix inversion (Rust): construct the matrix M = [t^{-1} - G0] for all sites within the FMS cluster (typically 50-200 atoms around absorber), invert to get full Green's function G = [t^{-1} - G0]^{-1}. Use LU decomposition or iterative Lanczos method for large clusters
  • 124.2.4.3 Implement XANES cross-section from FMS Green's function: mu(E) proportional to -Im[ sum_{L,L'} M*L(E) G{LL'}^{00}(E) M_{L'}(E) ], where M_L are dipole matrix elements between core state and final state angular momentum channel L, and G^{00} is the on-site (absorber) block of the full Green's function
  • 124.2.4.4 Implement dipole matrix elements: compute M_l = integral_0^{R_mt} R_core(r) _ r _ R_l(r) * r^2 dr for core state (1s for K-edge, 2p for L-edge) to final state angular momentum channels allowed by selection rules (Delta_l = +/-1 for dipole)
  • 124.2.4.5 Implement quadrupole matrix elements for pre-edge: M_{l,quad} = integral R_core(r) _ r^2 _ R_l(r) * r^2 dr for 1s->3d transitions at the K-edge of transition metals (Delta_l = +/-2 for quadrupole). Include electric quadrupole operator (k*r) coupling
  • 124.2.4.6 Implement cluster generation for FMS: given a crystal structure and absorber site, generate spherical cluster of specified radius (typically 6-8 Angstrom for XANES). Handle periodic boundary conditions, identify unique scattering paths

124.2.5 XANES — Core-Hole DFT Approaches#

  • 124.2.5.1 Implement full core-hole (FCH) approach: modify the pseudopotential/PAW dataset of the absorbing atom to remove one core electron, add one electron to the valence, perform self-consistent DFT calculation in a supercell. Extract unoccupied DOS projected onto absorber site as XANES proxy
  • 124.2.5.2 Implement half core-hole (HCH) / transition potential approach: Slater transition state with half a core electron removed. Compute eigenvalue differences as transition energies. This approximates the average of initial and final state effects
  • 124.2.5.3 Implement supercell convergence check: test XANES convergence with supercell size by computing spectra for 2x2x2, 3x3x3, etc. supercells. Core-hole interaction with periodic images must be negligible (energy shift < 0.1 eV between consecutive sizes)
  • 124.2.5.4 Implement core-hole screening analysis: compare FCH (fully screened), HCH (half screened), and unscreened (ground state DOS) to quantify screening effects on spectral shape and edge position

124.2.6 XANES — Bethe-Salpeter Equation (BSE) for Core Excitations#

  • 124.2.6.1 Implement core-level BSE Hamiltonian construction: build the electron-hole Hamiltonian H^{BSE}{vc,v'c'} = (E_c - E_v) delta{vv'} delta*{cc'} + K^{x}*{vc,v'c'} - K^{d}_{vc,v'c'} where v indexes core states and c indexes conduction states. Exchange kernel K^x from bare Coulomb, direct (screened) kernel K^d from W (Phase 118 GW)
  • 124.2.6.2 Implement screened core-hole interaction: compute the screened Coulomb interaction W(r,r') between core hole and excited electron using the dielectric function from RPA or GW. Use static screening approximation W(omega=0) for efficiency
  • 124.2.6.3 Implement BSE eigenvalue solver for core excitations: Lanczos/Haydock recursion method to compute the XAS spectrum without full diagonalization — only the spectral function Im[<phi_0|G(omega)|phi_0>] is needed where phi_0 is the dipole-weighted initial state
  • 124.2.6.4 Implement OCEAN-equivalent workflow: integrate GW quasiparticle energies (Phase 118), BSE kernel, and core-level matrix elements into a single pipeline. Validate against published OCEAN results for standard systems (diamond C K-edge, SrTiO3 Ti L-edge)

124.2.7 XANES — Edge Types and Special Effects#

  • 124.2.7.1 Implement K-edge XANES: 1s core hole, dipole selection rule 1s->p final states, quadrupole 1s->d pre-edge. Handle elements Z=3..92 with appropriate core-level energies from tabulated values
  • 124.2.7.2 Implement L-edge XANES: L1 (2s->p), L2 (2p1/2->s,d), L3 (2p3/2->s,d) edges. Include spin-orbit splitting of 2p core level. Handle branching ratio L3/L2 and statistical ratio deviation
  • 124.2.7.3 Implement M-edge XANES: M4,5 (3d->f) edges relevant for rare earths and actinides. Include multiplet effects from 3d-4f Coulomb and exchange interactions via atomic multiplet theory (Cowan code approach)
  • 124.2.7.4 Implement angular-dependent XANES (X-ray natural linear dichroism, XNLD): compute orientation-dependent absorption for anisotropic systems by resolving dipole matrix elements along crystal axes. Output sigma_x(E), sigma_y(E), sigma_z(E) for polarized measurements
  • 124.2.7.5 Implement X-ray magnetic circular dichroism (XMCD): compute difference spectrum mu*+(E) - mu*-(E) for left and right circularly polarized X-rays. Implement spin and orbital sum rules: m_orb = -integral(XMCD _ dE) _ 4q / (3*integral(XAS * dE)), m_spin = -integral(XMCD _ dE) _ (6p - 4q) / integral(XAS * dE) where p and q are L3 and L2+L3 integrated XMCD
  • 124.2.7.6 Implement pre-edge feature extraction: fit pre-edge region to Gaussian/Lorentzian peaks on a polynomial background, compute centroid energy and integrated intensity, relate to coordination geometry (tetrahedral vs. octahedral for transition metals via 1s->3d intensity)

124.2.8 EXAFS — Scattering Path Expansion#

  • 124.2.8.1 Implement EXAFS equation: chi(k) = sum_j (N_j _ S0^2 _ F_j(k)) / (k _ R_j^2) _ sin(2kR_j + phi_j(k)) * exp(-2sigma_j^2k^2) * exp(-2*R_j/lambda(k)) where the sum runs over scattering paths j, with degeneracy N_j, effective scattering amplitude F_j, phase shift phi_j, Debye-Waller factor sigma_j^2, and mean free path lambda(k)
  • 124.2.8.2 Implement scattering path enumeration: enumerate all single-scattering and multiple-scattering paths within a specified maximum half-path-length R_max (typically 6-8 Angstrom). Use heap-based path tree expansion with early termination for negligible amplitude paths
  • 124.2.8.3 Implement path importance filtering: compute curved-wave amplitude ratio for each path relative to the strongest single-scattering path. Filter paths by importance threshold (default: keep paths contributing > 2% of strongest path amplitude). Report path statistics
  • 124.2.8.4 Implement effective scattering amplitude F_eff(k) and phase phi_eff(k) for each path (Rust): for single scattering, F(k) = |f(theta=pi, k)| backscattering amplitude; for multiple scattering, compute the product of scattering matrices along the path using the plane-wave or curved-wave approximation
  • 124.2.8.5 Implement curved-wave multiple scattering for EXAFS paths (Rust): expand the scattering amplitude in partial waves, sum products of t-matrices and free propagators along each leg of the path. Handle up to 4-leg paths (single, double, triple, quadruple scattering)
  • 124.2.8.6 Implement mean free path lambda(k): energy-dependent photoelectron mean free path from the imaginary part of the optical potential. Use the empirical formula or compute from the self-energy

124.2.9 EXAFS — Debye-Waller Factors#

  • 124.2.9.1 Implement correlated Debye model for sigma^2: given a Debye temperature Theta_D, compute sigma^2(T) = (3*hbar^2) / (m_reduced * k_B _ Theta_D) _ [phi_1(Theta_D/T) / (Theta_D/T) + 1/4] where phi_1 is the first Debye function and m_reduced is the reduced mass of the absorber-scatterer pair
  • 124.2.9.2 Implement Einstein model for sigma^2: single-frequency approximation sigma^2(T) = hbar / (2muomega_E) * coth(hbaromega_E / (2k_B*T)) with Einstein frequency omega_E as fitting parameter
  • 124.2.9.3 Implement phonon-based Debye-Waller factors: compute sigma^2 from the phonon density of states projected onto the absorber-scatterer bond direction. Use dynamical matrix from Phase 119 (lattice dynamics) or AIMD trajectory correlation functions from Phase 120
  • 124.2.9.4 Implement third and fourth cumulants C3 and C4 for anharmonic EXAFS: C3 accounts for asymmetric pair distribution (skewness), C4 for kurtosis. Include these in the EXAFS equation as additional phase and amplitude corrections

124.2.10 EXAFS — Fitting Engine#

  • 124.2.10.1 Implement EXAFS fitting in k-space: minimize sum_i [k^w * chi_theory(k_i) - k^w * chi_exp(k_i)]^2 over structural parameters (R_j, sigma_j^2, N_j, delta_E0, S0^2). Support multiple k-weights w = 1, 2, 3 simultaneously
  • 124.2.10.2 Implement EXAFS fitting in R-space: Fourier transform both theory and experiment, fit in the specified R-range. Handle real and imaginary parts or magnitude and phase of chi(R)
  • 124.2.10.3 Implement Levenberg-Marquardt optimizer for EXAFS fitting: parameter constraints (bounds, fixed parameters, correlated parameters via math expressions), covariance matrix estimation, uncertainty propagation to structural parameters
  • 124.2.10.4 Implement EXAFS fit quality metrics: reduced chi-squared, R-factor = sum|chi_exp - chi_theory|^2 / sum|chi_exp|^2, number of independent data points N_idp = 2Delta_kDelta_R/pi, information criterion (AIC, BIC) for model comparison

124.2.11 XAS Spectrum Processing#

  • 124.2.11.1 Implement XAS background subtraction: pre-edge subtraction (linear or Victoreen fit), post-edge normalization (polynomial or spline fit to determine mu_0), edge step Delta_mu_0 determination. Follow IFEFFIT/Larch conventions
  • 124.2.11.2 Implement E0 determination: inflection point method (maximum of first derivative), half-edge-step method, or user-specified. Handle noisy data with smoothing before derivative
  • 124.2.11.3 Implement chi(k) extraction: subtract smooth background mu_0(E) using Autobk algorithm (spline with R_bkg parameter controlling low-R cutoff), convert to k-space via k = sqrt(2m(E-E0)/hbar^2)
  • 124.2.11.4 Implement Fourier transform of chi(k): FT with configurable window functions (Hanning, Kaiser-Bessel, Gaussian), k-range selection, k-weight, zero-padding. Output magnitude |chi(R)|, real part Re[chi(R)], imaginary part Im[chi(R)], phase
  • 124.2.11.5 Implement back-Fourier transform (Fourier filtering): select R-range window, back-transform to k-space to isolate contribution from specific coordination shells

124.3 X-ray Photoelectron Spectroscopy & Photoemission (@kalika/photoemission)#

124.3.1 Library Scaffolding#

  • 124.3.1.1 Create libs/kalika/photoemission/ directory with project.json, package.json, tsconfig.json. Add @kalika/photoemission path mapping to tsconfig.base.json

124.3.2 Core-Level Binding Energies#

  • 124.3.2.1 Implement initial-state (Koopmans) binding energy estimation: E_B approx -epsilon_KS where epsilon_KS is the Kohn-Sham eigenvalue of the core state. Include correction for the difference between KS eigenvalue and true ionization energy
  • 124.3.2.2 Implement Delta-SCF binding energy: compute E_B = E(N-1, core hole) - E(N, ground state) from two self-consistent DFT calculations — ground state and core-ionized state. Handle fractional occupation for convergence
  • 124.3.2.3 Implement Slater-Janak transition state: E_B = -epsilon(n=1/2) where the core orbital occupation is set to 1/2. Requires single SCF calculation, more efficient than full Delta-SCF
  • 124.3.2.4 Implement core-level shift (CLS) prediction between inequivalent sites: compute relative shifts Delta_E_B between different atomic environments in the same structure (e.g., surface vs. bulk atoms, different oxidation states). Initial state CLS from electrostatic potential differences, final state CLS from Delta-SCF differences

124.3.3 XPS Spectrum Simulation#

  • 124.3.3.1 Implement XPS chemical shift prediction: map computed core-level binding energies to experimental chemical shift scale using reference compound calibration (e.g., C 1s of adventitious carbon = 284.8 eV, Au 4f7/2 = 84.0 eV)
  • 124.3.3.2 Implement Auger parameter calculation: alpha' = E_K(KLL) + E_B(core) where E_K is the kinetic energy of the Auger line. The modified Auger parameter is instrument-independent and more reliable than binding energy alone for chemical state identification
  • 124.3.3.3 Implement surface core-level shifts: account for reduced coordination at surfaces by computing CLS for slab models with distinct surface, subsurface, and bulk-like layers. Include surface segregation effects for alloys
  • 124.3.3.4 Implement valence band XPS simulation: compute DOS(E) weighted by element-specific and orbital-specific photoionization cross-sections sigmanl(hv) at the given photon energy. Sum contributions from all atoms: I(E) = sum_i sum_l rho{i,l}(E) * sigma_{i,l}(hv)
  • 124.3.3.5 Implement satellite structure prediction: estimate shake-up (monopole excitation) and shake-off intensities from sudden approximation overlap integrals between N-1 electron states

124.3.4 Angle-Resolved Photoemission (ARPES) Simulation#

  • 124.3.4.1 Implement spectral function A(k,omega) from DFT bands: A(k,omega) = sumn delta(omega - epsilon{n,k}) broadened by Lorentzian (lifetime) and Gaussian (instrumental). Use band structure from Phase 116 DFT or Phase 118 GW
  • 124.3.4.2 Implement GW spectral function for ARPES: A(k,omega) = -1/pi * Im[Sigma(k,omega)] / [(omega - epsilon_k - Re[Sigma(k,omega)])^2 + Im[Sigma(k,omega)]^2] showing quasiparticle peaks with renormalized dispersion and satellites from GW self-energy (Phase 118)
  • 124.3.4.3 Implement photoemission matrix elements: compute |<f|A.p|i>|^2 where |i> is the initial Bloch state, A is the photon vector potential, p is the momentum operator, and |f> is the final photoelectron state. Include photon polarization dependence (s-pol, p-pol, circular)
  • 124.3.4.4 Implement free-electron final state approximation: model the final state as a plane wave |k_f> with k_f determined by energy conservation E_f = E_i + h*nu - phi. Map k_parallel from emission angle via k_parallel = sqrt(2mE_K)/hbar * sin(theta)
  • 124.3.4.5 Implement one-step model of photoemission: time-reversed LEED state as the final state, properly matching bulk and vacuum wave functions at the surface, including surface barrier effects
  • 124.3.4.6 Implement kz broadening for ARPES: account for finite photoelectron escape depth (surface sensitivity) which causes intrinsic kz uncertainty Delta_kz ~ 1/lambda, broadening features that disperse along kz. Convolve spectral function along kz with Lorentzian of width 1/lambda
  • 124.3.4.7 Implement resonant photoemission simulation: compute enhancement of valence band features when photon energy is tuned to a core-level absorption edge (Fano resonance profile). Calculate autoionization matrix elements

124.3.5 Spin-Resolved Photoemission#

  • 124.3.5.1 Implement spin-resolved ARPES: separate spectral function into spin-up and spin-down channels for magnetic systems. Compute spin polarization P(E,k) = [A_up - A_down] / [A_up + A_down]
  • 124.3.5.2 Implement spin-orbit induced spin texture: for topological surface states and Rashba systems, compute spin expectation value <S>(k) for each band from spinor wavefunctions. Display spin texture on Fermi surface

124.4 Vibrational Spectroscopy (@kalika/vibrational-spectroscopy)#

124.4.1 Library Scaffolding#

  • 124.4.1.1 Create libs/kalika/vibrational-spectroscopy/ directory with project.json, package.json, tsconfig.json. Add @kalika/vibrational-spectroscopy path mapping to tsconfig.base.json
  • 124.4.1.2 Create Rust crate for Raman tensor and polarizability derivative computations. Configure napi-rs bindings

124.4.2 Infrared (IR) Spectroscopy#

  • 124.4.2.1 Implement IR intensity from Born effective charges: IIR_s = |sum{kappa,alpha} Z*_{kappa,alpha,beta} * e_{kappa,alpha,s}|^2 summed over polarization direction beta, where Z* is the Born effective charge tensor, e is the phonon eigenvector for mode s, kappa is atom index, alpha is Cartesian direction. Use Z* from DFPT (Phase 119)
  • 124.4.2.2 Implement symmetry-based IR activity determination: a mode is IR-active if it belongs to an irreducible representation that transforms as a translation (x, y, or z) in the point group character table. Use site symmetry analysis from Phase 117 (crystal symmetry)
  • 124.4.2.3 Implement IR spectrum construction: place Lorentzian or Gaussian peaks at each IR-active phonon frequency with height proportional to computed IR intensity. Handle broadening parameter, temperature-dependent occupation via Bose-Einstein factor
  • 124.4.2.4 Implement peak assignment engine: for each IR peak, identify the phonon mode, compute atomic displacement amplitudes, classify by character (stretching, bending, rocking, wagging, twisting) using bond angle and bond length changes during vibration
  • 124.4.2.5 Implement anharmonic IR from AIMD: compute the IR spectrum from the autocorrelation of the total dipole moment time series from AIMD trajectories: I(omega) proportional to omega^2 _ integral <M(0).M(t)> _ exp(-iomegat) dt, where M(t) is the cell dipole at time t. Includes all anharmonic and temperature effects intrinsically
  • 124.4.2.6 Implement ATR-IR simulation: compute attenuated total reflectance spectrum from the complex dielectric function using the Fresnel equations for the crystal/ATR-element interface at the specified angle of incidence and ATR material (diamond, ZnSe, Ge)
  • 124.4.2.7 Implement reflectance IR and Kramers-Kronig analysis: compute reflectivity R(omega) = |(sqrt(epsilon) - 1)/(sqrt(epsilon) + 1)|^2 from the dielectric function. Implement Kramers-Kronig transform to extract optical constants n(omega), k(omega) from measured R(omega)

124.4.3 Raman Spectroscopy#

  • 124.4.3.1 Implement Raman tensor computation via finite differences (Rust): compute the derivative of the macroscopic dielectric tensor with respect to atomic displacements: alpha'{alpha,beta,s} = d(epsilon{alpha,beta})/d(Q_s) where Q_s is the normal mode coordinate. Use central differences: displace atoms +/- delta along each eigenvector, compute epsilon at each displacement via DFPT or finite-field DFT
  • 124.4.3.2 Implement DFPT Raman via 2n+1 theorem: compute the third-order energy derivative d^3E/(dE_i _ dE_j _ du_s) in one shot using density-functional perturbation theory, avoiding the 6*3N finite difference calculations. More efficient but requires implementation of third-order response
  • 124.4.3.3 Implement Raman intensity calculation: I_s proportional to |e_i . alpha'_s . e_s|^2 * (n_BE + 1) / omega_s for Stokes scattering, where e_i and e_s are incident and scattered polarization vectors, n_BE = 1/(exp(hbar*omega_s/k_B*T) - 1) is the Bose-Einstein factor, and omega_s is the phonon frequency
  • 124.4.3.4 Implement symmetry-based Raman activity determination: a mode is Raman-active if it belongs to an irreducible representation that transforms as a quadratic function (x^2, xy, xz, etc.) in the point group character table. Complementary to IR activity for centrosymmetric crystals (mutual exclusion rule)
  • 124.4.3.5 Implement polarization-dependent Raman: compute parallel (VV), perpendicular (VH), and depolarization ratio rho = I_VH / I_VV for each mode. For single crystals, compute the full angular dependence by rotating the Raman tensor with the crystal orientation matrix
  • 124.4.3.6 Implement resonant Raman spectroscopy: compute frequency-dependent polarizability tensor alpha(omega_L) near an electronic transition using sum-over-states expression with energy-dependent denominators. Identify modes enhanced by resonance (Albrecht A-term and B-term mechanisms)
  • 124.4.3.7 Implement temperature-dependent Raman: Stokes/anti-Stokes intensity ratio I_AS/I_S = (omega_L + omega_s)^4 / (omega_L - omega_s)^4 * exp(-hbar*omega_s/(k_B*T)) for temperature measurement from Raman. Include frequency shifts from anharmonicity
  • 124.4.3.8 Implement SERS enhancement factor estimation: compute electromagnetic enhancement |E_local/E_0|^4 from the local field enhancement at a plasmonic nanoparticle surface (Mie theory for spheres, FDTD/BEM reference data for other shapes). Estimate total enhancement combining electromagnetic and chemical (charge transfer) mechanisms
  • 124.4.3.9 Implement Raman for 2D materials: identify layer breathing modes (out-of-plane, frequency depends on number of layers via linear chain model) and shear modes (in-plane rigid layer sliding). Compute layer-number-dependent frequencies and use as thickness fingerprint

124.4.4 Phonon Mode Visualization and Group Theory#

  • 124.4.4.1 Implement phonon displacement vector visualization: for each phonon mode, generate a 3D representation of atomic displacement vectors (arrows on atoms). Export as JSON for WebGL rendering or as XYZ animation frames
  • 124.4.4.2 Implement group theory analysis of phonon modes at Gamma: decompose the 3N-dimensional representation of atomic displacements into irreducible representations of the crystal point group. Assign each mode its irrep label (e.g., A1g, E2g, T1u). Identify acoustic modes (T1u for cubic) and optical modes
  • 124.4.4.3 Implement selection rule summary: for each irrep, determine IR activity (transforms as x,y,z), Raman activity (transforms as quadratic), hyper-Raman activity. Generate a selection rule table compatible with standard spectroscopy references

124.5 Nuclear Magnetic Resonance (@kalika/nmr)#

124.5.1 Library Scaffolding#

  • 124.5.1.1 Create libs/kalika/nmr/ directory with project.json, package.json, tsconfig.json. Add @kalika/nmr path mapping to tsconfig.base.json
  • 124.5.1.2 Create Rust crate for GIPAW kernel computations (current-density response, gauge-origin handling). Configure napi-rs

124.5.2 Chemical Shielding Tensor (GIPAW Method)#

  • 124.5.2.1 Implement GIPAW chemical shielding tensor calculation: sigma*{alpha,beta} = d(B_induced*{alpha}) / d(Bext{beta}) computed as the linear response of the induced magnetic field at the nuclear site to a uniform external field. GIPAW (Gauge-Including Projector Augmented Waves) handles the gauge-origin problem for PAW pseudopotentials by including the all-electron current density within augmentation spheres
  • 124.5.2.2 Implement bare (pseudized) contribution to shielding: the current density from smooth pseudo-wavefunctions jps(r) responding to uniform B field via first-order perturbation theory: |psi_n^(1)> = sum{m!=n} |psi_m^(0)><psi_m^(0)|H^(1)|psi_n^(0)> / (E_n - E_m) where H^(1) = (1/2c)(r x p).B + (B x r).p is the perturbation from the external magnetic field
  • 124.5.2.3 Implement augmentation (PAW reconstruction) contribution: reconstruct the all-electron current density within PAW spheres from the pseudo-wavefunctions, accounting for the full nodal structure of core and semicore states. This correction is critical for accurate absolute shieldings
  • 124.5.2.4 Implement isotropic chemical shift extraction: delta_iso = sigma_ref - sigma_iso where sigma_iso = (sigma_xx + sigma_yy + sigma_zz)/3 is the isotropic shielding and sigma_ref is the reference compound shielding. Compute chemical shift anisotropy (CSA): Delta_sigma = sigma_zz - (sigma_xx + sigma_yy)/2 and asymmetry parameter eta = (sigma_yy - sigma_xx) / (sigma_zz - sigma_iso) with |sigma_zz - sigma_iso| >= |sigma_xx - sigma_iso| >= |sigma_yy - sigma_iso|
  • 124.5.2.5 Implement shielding tensor eigenvalue decomposition: diagonalize the symmetric part of sigma to get principal components (sigma_11, sigma_22, sigma_33) and principal axis orientations. Visualize as shielding tensor ellipsoid on atomic site

124.5.3 Electric Field Gradient (EFG) Tensor#

  • 124.5.3.1 Implement EFG tensor computation: V_{alpha,beta} = d^2(V_Coulomb)/d(r_alpha)*d(r_beta) evaluated at each nuclear site, where V_Coulomb is the total electrostatic potential from electrons and nuclei. Subtract the trace (Laplacian = 0 at nuclear site by Poisson equation for point charges)
  • 124.5.3.2 Implement PAW reconstruction for EFG: the all-electron charge density near the nucleus determines the EFG. Reconstruct from PAW: V*{ae} = V*{ps} + sum*{ij} (rho^1*{ij} <phii|V_op|phi_j> - rho^1{ij} <phi~_i|V_op|phi~_j>) where phi and phi~ are AE and PS partial waves
  • 124.5.3.3 Implement quadrupolar coupling parameters: compute C_Q = eQV_zz/h (quadrupolar coupling constant, in MHz) and eta_Q = (V_xx - V_yy)/V_zz (asymmetry parameter, 0 <= eta <= 1) for each nuclear site with quadrupole moment Q > 0. Tabulate Q values for all quadrupolar nuclei

124.5.4 J-Coupling and Other NMR Parameters#

  • 124.5.4.1 Implement indirect nuclear spin-spin coupling (J-coupling) from DFT: compute the four contributions — Fermi contact (FC), spin-dipole (SD), paramagnetic spin-orbit (PSO), diamagnetic spin-orbit (DSO) — using linear response to a nuclear magnetic dipole perturbation. The Fermi contact term dominates for light nuclei
  • 124.5.4.2 Implement Knight shift for metallic systems: K = K_spin + K_orbital + K_dia where K_spin = (8pi/3) _ chi_Pauli _ |psi(0)|^2 / rho is the contact Knight shift from the spin susceptibility and the electron density at the nucleus
  • 124.5.4.3 Implement paramagnetic NMR shifts: Fermi contact shift delta_FC proportional to A_iso _ <S_z> / (gamma_N _ B0) where A_iso is the isotropic hyperfine coupling and <S_z> is the thermal average of electronic spin. Pseudocontact shift (dipolar) delta_PCS from the anisotropic magnetic susceptibility tensor

124.5.5 NMR Spectrum Simulation#

  • 124.5.5.1 Implement powder-averaged NMR spectrum: for each nuclear site, convolve the CSA powder pattern (from shielding tensor eigenvalues) with quadrupolar splitting pattern (for I > 1/2 nuclei). Use ZCW (Zaremba-Conroy-Wolfsberg) angular grid for efficient powder averaging over orientations
  • 124.5.5.2 Implement MAS (Magic Angle Spinning) NMR simulation: compute spinning sideband pattern using Herzfeld-Berger analysis — sideband intensities from ratios of CSA principal values at given spinning rate. Include second-order quadrupolar broadening for half-integer quadrupolar nuclei under MAS
  • 124.5.5.3 Implement 2D NMR prediction — HETCOR (heteronuclear correlation): compute cross-peaks between pairs of nuclei (e.g., 1H-13C) based on spatial proximity (dipolar coupling strength proportional to 1/r^3). Filter by distance cutoff
  • 124.5.5.4 Implement 2D MQMAS (multiple-quantum MAS) simulation: for half-integer quadrupolar nuclei, compute isotropic projection (free from second-order quadrupolar broadening) vs. MAS dimension. Resolve inequivalent sites by their isotropic chemical shift and quadrupolar parameters

124.5.6 Reference Compounds and Common Nuclei#

  • 124.5.6.1 Implement reference compound database: store computed absolute shieldings for standard NMR reference compounds — TMS (tetramethylsilane) for 1H/13C/29Si, liquid NH3 for 15N, 85% H3PO4 for 31P, CFCl3 for 19F, NaCl(aq) for 23Na, Al(H2O)6^3+ for 27Al, etc. Include basis set / functional / method dependence
  • 124.5.6.2 Implement chemical shift scale calibration: given the computed absolute shielding sigma_calc and reference shielding sigma_ref, convert to chemical shift delta = sigma_ref - sigma_calc. Provide empirical linear regression correction: delta_corr = a * delta_calc + b with a, b fitted to benchmark sets for each nucleus/functional combination

124.6 Mossbauer Spectroscopy (@kalika/mossbauer)#

124.6.1 Library Scaffolding#

  • 124.6.1.1 Create libs/kalika/mossbauer/ directory with project.json, package.json, tsconfig.json. Add @kalika/mossbauer path mapping to tsconfig.base.json

124.6.2 Isomer Shift#

  • 124.6.2.1 Implement isomer shift from contact density: delta = alpha * [rho(0)_absorber - rho(0)_source] where rho(0) is the total electron contact density at the Mossbauer nucleus. Compute rho(0) from the all-electron charge density at the nuclear position using PAW reconstruction. The calibration constant alpha depends on the nuclear transition (e.g., alpha = -0.291 mm/s / a0^-3 for 57Fe)
  • 124.6.2.2 Implement calibration constant database: store alpha values for common Mossbauer isotopes — 57Fe (14.4 keV), 119Sn (23.9 keV), 151Eu (21.5 keV), 161Dy (25.7 keV), 197Au (77.3 keV). Include reference source rho(0) values (alpha-Fe for 57Fe, BaSnO3 for 119Sn)
  • 124.6.2.3 Implement isomer shift validation suite: compute isomer shifts for a test set of Fe compounds with known experimental values (Fe metal, FeO, Fe2O3, Fe3O4, FeS2, FeCO3) and verify agreement within +/- 0.1 mm/s using PBE functional

124.6.3 Quadrupole Splitting#

  • 124.6.3.1 Implement quadrupole splitting from EFG: Delta_E_Q = (1/2) _ e _ Q _ V_zz _ sqrt(1 + eta^2/3) where V_zz is the largest eigenvalue of the EFG tensor, Q is the nuclear quadrupole moment, and eta is the asymmetry parameter. Reuse EFG computation from NMR module (124.5.3)
  • 124.6.3.2 Implement quadrupole splitting for I=3/2 excited state (57Fe): the excited state (I=3/2) splits into m_I = +/-3/2 and +/-1/2 sublevels with separation Delta_E_Q. The ground state (I=1/2) is unsplit. Produce two-line pattern with splitting Delta_E_Q and equal intensity ratio (3:2:1:1:2:3 ratios for magnetic splitting)

124.6.4 Magnetic Hyperfine Field#

  • 124.6.4.1 Implement magnetic hyperfine field B_hf: compute three contributions — (1) Fermi contact B_FC = -(8pi/3) _ mu_B _ [rho_up(0) - rho_down(0)] from spin-polarized contact density, (2) dipolar B_dip from spin-density distribution within atomic sphere, (3) orbital B_orb from orbital angular momentum. Total B_hf = B_FC + B_dip + B_orb
  • 124.6.4.2 Implement core polarization contribution to B_FC: the dominant contribution comes from spin polarization of core s-electrons (1s, 2s, 3s for Fe) by the exchange interaction with unpaired d-electrons. Requires all-electron or PAW reconstruction to capture core spin density
  • 124.6.4.3 Implement six-line magnetic Mossbauer pattern: for magnetically ordered systems, the six allowed transitions between I=1/2 (m = +/-1/2) and I=3/2 (m = +/-3/2, +/-1/2) states produce a sextet with velocity positions determined by B_hf and relative intensities 3:x:1:1:x:3 where x depends on the angle between B_hf and gamma-ray direction (x=2 for powder, x=0 for B parallel to gamma)

124.6.5 Mossbauer Spectrum Simulation#

  • 124.6.5.1 Implement Mossbauer spectrum line shape: each transition is a Lorentzian with natural linewidth Gamma_0 (for 57Fe, Gamma_0 = 0.097 mm/s = 4.67 neV). Total spectrum is sum of Lorentzians: I(v) = sum_i A_i * (Gamma/2)^2 / [(v - v_i)^2 + (Gamma/2)^2] where v_i are transition velocities and A_i are relative intensities
  • 124.6.5.2 Implement combined isomer shift + quadrupole + magnetic spectrum: handle the general case with all three hyperfine interactions present simultaneously. Diagonalize the full nuclear Hamiltonian H = H_M + H_Q including off-diagonal quadrupolar terms when the EFG principal axes do not coincide with the magnetic field direction
  • 124.6.5.3 Implement Lamb-Mossbauer factor (recoil-free fraction): f = exp(-k^2 * <u^2>) where k is the gamma-ray wavevector and <u^2> is the mean-square displacement of the Mossbauer nucleus. Compute <u^2> from phonon calculations (Phase 119) or Debye model: <u^2> = (3hbar^2T)/(Mk_BTheta_D^2) * phi_1(Theta_D/T) + 3hbar^2/(4Mk_BTheta_D)
  • 124.6.5.4 Implement oxidation state and spin state identification: from computed isomer shift delta and quadrupole splitting Delta_E_Q, classify the iron site: Fe(0) metallic, Fe(II) HS/LS, Fe(III) HS/LS, Fe(IV), Fe(VI). Use decision boundaries from empirical Mossbauer parameter ranges (e.g., Fe(II) HS: delta ~ 0.9-1.4 mm/s, Fe(III) HS: delta ~ 0.3-0.5 mm/s)

124.7 Electron Energy Loss Spectroscopy (@kalika/eels)#

124.7.1 Library Scaffolding#

  • 124.7.1.1 Create libs/kalika/eels/ directory with project.json, package.json, tsconfig.json. Add @kalika/eels path mapping to tsconfig.base.json
  • 124.7.1.2 Create Rust crate for EELS cross-section and energy loss function computations. Configure napi-rs bindings

124.7.2 Low-Loss EELS#

  • 124.7.2.1 Implement energy loss function: compute -Im[1/epsilon(q,omega)] from the macroscopic dielectric function epsilon(q,omega) obtained from TDDFT or GW/BSE (Phase 118). The loss function peaks at collective excitations (plasmons) and interband transitions
  • 124.7.2.2 Implement plasmon energy and dispersion: locate the energy where Re[epsilon(q,omega)] = 0 and Im[epsilon] is small — this defines the bulk plasmon energy omega_p(q). For free-electron-like metals, omega_p = sqrt(4pin*e^2/m). Track plasmon dispersion omega_p(q) along high-symmetry directions
  • 124.7.2.3 Implement q-dependent dielectric function: compute epsilon(q,omega) beyond the optical limit (q->0) by including local field effects and finite momentum transfer. Use the density-density response function chi(q,omega) from TDDFT: epsilon^{-1}(q,omega) = 1 + v(q) * chi(q,omega)
  • 124.7.2.4 Implement interband transition identification: decompose the loss function into contributions from specific band-to-band transitions by computing the joint density of states weighted by matrix elements. Label each feature with the contributing bands
  • 124.7.2.5 Implement Kramers-Kronig analysis for EELS: extract the real and imaginary parts of epsilon from the loss function using KK relations: Re[1/epsilon] = 1 - (2/pi) _ P integral [omega' _ Im[-1/epsilon(omega')]] / (omega'^2 - omega^2) d(omega'). Validate by checking f-sum rule: integral omega _ Im[-1/epsilon] d(omega) = pi/2 _ omega_p^2
  • 124.7.2.6 Implement Cerenkov radiation loss subtraction: for relativistic electrons (>100 keV beam), compute and subtract the Cerenkov loss contribution that appears when electron velocity exceeds the phase velocity of light in the material (v > c/sqrt(epsilon)). Relevant for low-loss EELS of semiconductors

124.7.3 Core-Loss EELS#

  • 124.7.3.1 Implement core-loss EELS edge structure (ELNES): compute the energy-loss near-edge structure equivalent to XANES using the double-differential cross-section d^2sigma/(dEdOmega) = (4gamma^2 / a0^2*q^4) * S(q,E) where S(q,E) is the dynamic structure factor and gamma is the relativistic factor. For small q, ELNES converges to XANES (dipole limit)
  • 124.7.3.2 Implement EXELFS (extended energy loss fine structure): extract oscillatory fine structure beyond the edge onset, analogous to EXAFS. Use same scattering path formalism with appropriate scattering geometry corrections for the EELS momentum transfer
  • 124.7.3.3 Implement collection/convergence angle effects: integrate the double-differential cross-section over the experimental collection semi-angle beta and convergence semi-angle alpha. This modifies the relative weights of dipole and non-dipole transitions, especially important for L- and M-edges
  • 124.7.3.4 Implement generalized oscillator strength (GOS) database: tabulate or compute GOS f(q,E) = (E _ q^2) / (2 _ R) * |<f|exp(iq.r)|i>|^2 for all edges of elements Z=3..92. Store on a (q,E) grid for interpolation during cross-section calculation
  • 124.7.3.5 Implement white line ratio analysis: for L2,3 edges of transition metals, compute the L3/L2 intensity ratio (integrated over the white line region). Deviation from the statistical ratio (2:1) indicates d-electron count and spin-orbit coupling. Implement the relationship between L3/L2 ratio and d-band occupancy
  • 124.7.3.6 Implement EELS quantification: compute absolute partial cross-sections sigma_k(beta, Delta) for each edge (integrated over energy window Delta and collection angle beta). Estimate composition from N_A/N_B = (I_A _ sigma_B) / (I_B _ sigma_A) using the k-factor method

124.7.4 Momentum-Resolved and Vibrational EELS#

  • 124.7.4.1 Implement momentum-resolved EELS: compute the dynamic structure factor S(q,omega) from TDDFT as a 2D function of momentum transfer q and energy loss omega. Display as an intensity map showing plasmon dispersion, particle-hole continuum, and excitonic features
  • 124.7.4.2 Implement vibrational EELS (high-resolution EELS): compute the dipole scattering cross-section for surface and bulk phonon losses at meV-scale resolution. Include impact scattering mechanism for dipole-forbidden modes at off-specular angles. Interface with phonon calculations from Phase 119

124.8 Other Spectroscopies (@kalika/spectroscopy-extended)#

124.8.1 Library Scaffolding#

  • 124.8.1.1 Create libs/kalika/spectroscopy-extended/ directory with project.json, package.json, tsconfig.json. Add @kalika/spectroscopy-extended path mapping to tsconfig.base.json

124.8.2 Optical Absorption and UV-Vis#

  • 124.8.2.1 Implement UV-Vis absorption spectrum from dielectric function: alpha(omega) = omega _ Im[sqrt(epsilon(omega))] / c, or for thin films, absorbance A = alpha _ d. Import epsilon(omega) from Phase 123 (optical properties) or compute from TDDFT/BSE (Phase 118)
  • 124.8.2.2 Implement oscillator strength and transition dipole moments: for each optical transition, compute f_n = (2momega_n)/(3*hbar) * |<n|r|0>|^2. Sum rule: sum_n f_n = N_electrons. Relate to molar extinction coefficient epsilon_M via integration over absorption band

124.8.3 Photoluminescence (PL)#

  • 124.8.3.1 Implement PL spectrum from configuration coordinate (CC) model: given the CC diagram from Phase 122 (defect calculations), compute the emission spectrum as a Poisson distribution of phonon sidebands: I(E) = sum_n [S^n * exp(-S) / n!] * L(E - E_ZPL + n*hbar*omega) where S is the Huang-Rhys factor, E_ZPL is the zero-phonon line energy, and omega is the effective phonon frequency
  • 124.8.3.2 Implement Huang-Rhys factor computation: S = Delta_Q^2 * omega / (2*hbar) where Delta_Q is the displacement of the equilibrium position between ground and excited state potential energy surfaces along the generalized configuration coordinate. Compute from Delta_Q = sqrt(sum_i m_i * (R_i^{excited} - R_i^{ground})^2)
  • 124.8.3.3 Implement zero-phonon line (ZPL) energy: E_ZPL = E_abs - E_relax_excited - E_relax_ground (vertical absorption minus relaxation energies in both states). The ZPL is the sharp feature in low-temperature PL spectra
  • 124.8.3.4 Implement Stokes shift: Delta_Stokes = E_abs - E_em = 2 _ S _ hbar * omega. Compute from CC diagram as the energy difference between vertical absorption and vertical emission
  • 124.8.3.5 Implement radiative recombination rate: compute the radiative lifetime tau_rad = 1 / (n _ e^2 _ omega^2 * |<f|r|i>|^2 / (3piepsilon_0mc^3*hbar)) from the transition dipole matrix element between defect states. Compare with experimental PL decay times
  • 124.8.3.6 Implement temperature-dependent PL: broaden phonon sidebands with thermal occupation, include non-radiative decay via multi-phonon capture (exponential gap law). Simulate PL spectra at different temperatures

124.8.4 X-ray Emission Spectroscopy (XES)#

  • 124.8.4.1 Implement core X-ray emission lines: Kalpha (2p->1s) and Kbeta (3p->1s) emission energies and intensities. Kalpha1,2 splitting from 2p spin-orbit. Compute from the energy differences between core levels and transition dipole matrix elements
  • 124.8.4.2 Implement valence-to-core (VtC) XES: emission from occupied valence states filling the core hole. VtC-XES probes the occupied electronic structure (complementary to XANES which probes unoccupied). Compute from transition matrix elements <valence|r|1s> weighted by the valence DOS
  • 124.8.4.3 Implement Kbeta mainline analysis: the Kbeta1,3 and Kbeta' splitting in transition metals is sensitive to spin state (exchange interaction between 3p core hole and 3d valence). Compute the 3p-3d exchange integral and multiplet splitting

124.8.5 Resonant Inelastic X-ray Scattering (RIXS)#

  • 124.8.5.1 Implement RIXS cross-section (Kramers-Heisenberg formula): I(omega_in, omega_out) proportional to |sum_n <f|T|n><n|T|i> / (E_i + hbaromega_in - E_n + iGamma_n)|^2 where T is the transition operator, |n> are intermediate (core-excited) states, and the sum runs over all intermediate states
  • 124.8.5.2 Implement RIXS map computation: compute 2D intensity map I(excitation_energy, energy_loss) by scanning the incident photon energy across an absorption edge. Identify dd excitations (crystal field transitions in d-electron systems), charge transfer excitations, and magnon excitations as features in the RIXS map
  • 124.8.5.3 Implement dd excitations for transition metals: compute crystal field excitation energies and RIXS intensities using atomic multiplet theory (Tanabe-Sugano diagrams) with crystal field parameters 10Dq from DFT. Handle octahedral, tetrahedral, and lower-symmetry environments

124.8.6 Electron Paramagnetic Resonance (EPR)#

  • 124.8.6.1 Implement g-tensor from DFT+SOC: g*{ij} = g_e * delta*{ij} + 2 * sum_{n,occ; m,virt} <psi_n|L_i + g_e*S_i|psi_m> _ <psi_m|xi _ L_j|psi_n> / (E_n - E_m) where xi is the spin-orbit coupling constant, L is orbital angular momentum, S is spin, and the sum runs over occupied-virtual pairs. Compute principal g-values (g_xx, g_yy, g_zz) and principal axis orientations
  • 124.8.6.2 Implement hyperfine coupling constants: isotropic Aiso = (8pi/3) * g_e * mu_B * g_N * mu_N * rho_spin(R_N) (Fermi contact from spin density at nucleus), anisotropic A_aniso from the traceless dipolar tensor T*{ij} = muB * g_N * mu_N * integral rho_spin(r) * (3r_ir_j - r^2*delta{ij}) / r^5 dV
  • 124.8.6.3 Implement zero-field splitting (ZFS) parameters: for S >= 1 systems, compute the ZFS tensor D from spin-orbit and spin-spin coupling. Extract axial (D) and rhombic (E) parameters: D = D_zz - (D_xx + D_yy)/2, E = (D_xx - D_yy)/2
  • 124.8.6.4 Implement EPR spectrum simulation: given g-tensor, A-tensor(s), and D-tensor, simulate the EPR spectrum by diagonalizing the spin Hamiltonian H = mu_B * B . g . S + S . D . S + sum_k S . A_k . I_k at each field value. Handle powder averaging for polycrystalline samples, single-crystal angular dependence

124.8.7 Compton Scattering and Positron Annihilation#

  • 124.8.7.1 Implement Compton profile J(p_z): compute the electron momentum density rho(p) by Fourier transforming the Bloch wavefunctions to momentum space, then integrate over p_x and p_y to obtain the 1D Compton profile J(p_z) = integral integral rho(p) dp_x dp_y. Compare with experimental Compton scattering measurements
  • 124.8.7.2 Implement positron annihilation lifetime: solve the Schrodinger equation for the positron in the crystal potential (attractive at electron-rich regions, repulsive at nuclei). Compute the annihilation rate lambda = pi _ r_e^2 _ c _ integral n_+(r) _ n*-(r) * gamma(n*-(r)) dr where n*+ is positron density, n*- is electron density, and gamma is the enhancement factor. Lifetime tau = 1/lambda. Sensitive to vacancy defects (longer lifetime)

124.9 Spectrum Processing & Comparison (@kalika/spectrum-processing)#

124.9.1 Library Scaffolding#

  • 124.9.1.1 Create libs/kalika/spectrum-processing/ directory with project.json, package.json, tsconfig.json. Add @kalika/spectrum-processing path mapping to tsconfig.base.json

124.9.2 Peak Fitting#

  • 124.9.2.1 Implement Gaussian peak function: G(x) = A * exp(-(x - x0)^2 / (2sigma^2)) with amplitude A, center x0, and width sigma. FWHM = 2sqrt(2ln(2))sigma. Compute area = Asigmasqrt(2*pi)
  • 124.9.2.2 Implement Lorentzian peak function: L(x) = A * gamma^2 / ((x - x0)^2 + gamma^2) with amplitude A, center x0, and half-width gamma. FWHM = 2gamma. Compute area = Api*gamma
  • 124.9.2.3 Implement Voigt peak function: convolution of Gaussian and Lorentzian. Use Faddeeva function (complex error function) for efficient computation: V(x) = Re[w(z)] / (sigmasqrt(2pi)) where z = ((x-x0) + igamma) / (sigmasqrt(2)). Use the implementations in the Rust kernel for speed
  • 124.9.2.4 Implement pseudo-Voigt approximation: pV(x) = eta _ L(x) + (1-eta) _ G(x) where eta is the mixing parameter (0 = pure Gaussian, 1 = pure Lorentzian). Approximate the Voigt FWHM relation: f_V^5 = f_G^5 + 2.69f_G^4f_L + 2.43f_G^3f_L^2 + 4.47f_G^2f_L^3 + 0.07f_Gf_L^4 + f_L^5 (Thompson-Cox-Hastings formula)
  • 124.9.2.5 Implement Doniach-Sunjic line shape: DS(x) = cos(pi*alpha/2 + (1-alpha)*arctan((x-x0)/gamma)) / ((x-x0)^2 + gamma^2)^((1-alpha)/2) for metallic XPS core levels with asymmetry parameter alpha (0 = symmetric Lorentzian). Relevant for metals where electron-hole pair excitations create asymmetric tails
  • 124.9.2.6 Implement multi-peak fitting engine: given a spectrum and initial peak guesses (positions, widths, shapes), minimize chi^2 = sum_i (y_data_i - y_model_i)^2 / sigma_i^2 using Levenberg-Marquardt optimization. Support parameter constraints (fixed, bounded, linked between peaks). Report fit quality (chi^2_red, R^2), parameter uncertainties from covariance matrix, and residual spectrum
  • 124.9.2.7 Implement automatic peak detection: identify peaks using second derivative method (minima of d^2y/dx^2) or continuous wavelet transform (CWT). Provide initial guesses for multi-peak fitting. Handle noise with smoothing pre-filter

124.9.3 Background Subtraction#

  • 124.9.3.1 Implement linear background: fit y = a*x + b to user-specified background regions (before and after spectral features). Subtract from spectrum
  • 124.9.3.2 Implement Shirley background for XPS: iterative algorithm Bn(E) = k * integral{E}^{Emax} [I(E') - B{n-1}(E')] dE' with k set by matching endpoint intensities. Iterate until convergence (< 0.1% change). Accounts for inelastically scattered photoelectrons
  • 124.9.3.3 Implement Tougaard background for XPS: B(E) = lambda _ integral_{E}^{infinity} K(E'-E) _ [I(E') - B(E')] dE' using the three-parameter universal cross-section K(T) = B*T / (C + T^2)^2. More physically motivated than Shirley, applicable over wider energy range

124.9.4 Smoothing, Deconvolution, and Calibration#

  • 124.9.4.1 Implement Savitzky-Golay smoothing: fit polynomial of degree p to a window of 2m+1 points, use the polynomial value at the center as the smoothed value. Compute convolution coefficients analytically. Support derivative computation (1st, 2nd, 3rd order) simultaneously with smoothing
  • 124.9.4.2 Implement spectral deconvolution: given an instrument response function (typically Gaussian), deconvolve the measured spectrum to recover the intrinsic spectrum. Use Richardson-Lucy iterative deconvolution or Wiener filter with regularization to handle noise amplification
  • 124.9.4.3 Implement energy calibration and alignment: shift and stretch the energy axis to align features with known reference positions. Linear correction (offset + scale) or polynomial for non-linear detector response. For XPS: align to adventitious C 1s = 284.8 eV

124.9.5 Broadening and Convolution#

  • 124.9.5.1 Implement Gaussian broadening: convolve spectrum with Gaussian of specified FWHM (instrumental resolution). Use FFT-based convolution for efficiency on large spectra
  • 124.9.5.2 Implement Lorentzian broadening: convolve with Lorentzian (core-hole lifetime broadening). Support energy-dependent broadening width Gamma(E) that increases with energy above threshold (as final state lifetime decreases)
  • 124.9.5.3 Implement Voigt broadening: sequential Gaussian and Lorentzian convolution, or direct Voigt convolution using FFT. Standard for comparison of computed spectra with experiment

124.9.6 Spectrum Import and Comparison#

  • 124.9.6.1 Implement spectrum file parser: read common spectral data formats — multi-column ASCII (energy, intensity), XDI (X-ray Data Interchange for XAS), VAMAS (ISO 14976 for XPS), JCAMP-DX (spectroscopy interchange), and Athena project files (.prj). Auto-detect format from header
  • 124.9.6.2 Implement theory-experiment comparison metrics: R-factor R = sum|y_exp - y_theory|^2 / sum|y_exp|^2, chi-squared chi^2 = sum((y_exp - y_theory)/sigma)^2, cosine similarity cos_sim = (y_exp . y_theory) / (|y_exp| * |y_theory|), and Pearson correlation coefficient. Report all metrics for each comparison
  • 124.9.6.3 Implement spectrum alignment for comparison: cross- correlation based energy shift detection between theory and experiment. Optimize shift and scale factor to minimize R-factor. Handle different energy grids via interpolation (cubic spline)
  • 124.9.6.4 Implement reference spectrum database: store computed spectra for common compounds indexed by structure, spectroscopy type, computational method, and edge/nucleus. Enable lookup by chemical formula, space group, or spectral fingerprint similarity

124.10 Spectroscopy Visualization (@kalika/spectroscopy-viz)#

124.10.1 Library Scaffolding#

  • 124.10.1.1 Create libs/kalika/spectroscopy-viz/ directory with project.json, package.json, tsconfig.json. Add @kalika/spectroscopy-viz path mapping to tsconfig.base.json

124.10.2 1D Spectrum Plots#

  • 124.10.2.1 Implement interactive spectrum plot renderer: render spectrum as a line plot with energy/frequency on x-axis and intensity on y-axis. Support zoom (scroll wheel), pan (drag), cursor readout (crosshair with x,y values), and peak labels (position, assignment). Output as SVG for publication and as JSON scene graph for WebGL
  • 124.10.2.2 Implement multi-spectrum overlay: plot multiple spectra on the same axes with different colors/line styles. Support vertical offset (stacked waterfall plot) for comparing spectra at different conditions (temperature, pressure, composition)
  • 124.10.2.3 Implement spectral decomposition visualization: display individual peak components (from peak fitting) as filled areas beneath the total fit curve. Show residual (data - fit) in a separate panel below the main spectrum
  • 124.10.2.4 Implement theory vs. experiment comparison plot: overlay computed spectrum (broadened, aligned) with experimental data points (with error bars if available). Display R-factor and other comparison metrics in the legend. Highlight spectral regions with largest discrepancy

124.10.3 2D Spectrum Maps#

  • 124.10.3.1 Implement RIXS map visualization: render 2D intensity map with incident energy on one axis and energy loss (or emission energy) on the other. Use colormap (viridis, plasma, or custom) with logarithmic intensity scaling option. Mark constant emission energy (fluorescence) and constant energy loss (Raman) lines
  • 124.10.3.2 Implement ARPES visualization: render A(k,omega) as 2D intensity map with crystal momentum k on x-axis (along high-symmetry path) and binding energy on y-axis. Overlay DFT band structure as lines. Support multiple Brillouin zone paths in a single plot with high-symmetry point labels
  • 124.10.3.3 Implement 2D NMR correlation spectrum plot: render cross-peak intensity map for HETCOR/MQMAS spectra with chemical shift on both axes. Include contour lines and projections along each axis

124.10.4 Decomposed and Projected Spectra#

  • 124.10.4.1 Implement orbital-decomposed spectral contribution plot: for XAS, EELS, and XPS, show contributions from different angular momentum channels (s, p, d, f) as stacked areas or separate traces. Color-coded by orbital character
  • 124.10.4.2 Implement atom-projected spectral weight: for a multi-site structure, show contributions from each inequivalent atomic site to the total spectrum. Enable toggling individual site contributions on/off
  • 124.10.4.3 Implement band-decomposed optical spectrum: for UV-Vis and EELS, show which band-to-band transitions contribute to each spectral feature. Display as a heatmap of transition contributions indexed by (initial band, final band, energy)

124.10.5 Real-Space Visualization#

  • 124.10.5.1 Implement core orbital visualization: render the spatial distribution of the core state involved in XAS/XPS transitions (1s, 2p, 3d orbital shape) as isosurface on the atomic site. Use standard orbital shapes with radial function from the atomic calculation
  • 124.10.5.2 Implement excited-state (exciton) wavefunction visualization: for BSE-computed core excitons, display the electron density distribution |psi_e(r)|^2 with the hole fixed at the absorber site. Shows the spatial extent and character (localized vs. delocalized) of the X-ray excited state as a 3D isosurface on the crystal structure
  • 124.10.5.3 Implement NMR shielding tensor ellipsoid visualization: display the chemical shielding tensor as an ellipsoid at each nuclear site, with principal axes and eigenvalues shown. Orientation relative to crystal axes reveals bonding symmetry

124.11 Cross-Module Integration & Validation (@kalika/spectroscopy)#

124.11.1 Phases 116 & 118 Integration (Electronic Structure)#

  • 124.11.1.1 Implement DFT wavefunction interface: read Kohn-Sham wavefunctions, eigenvalues, and density from the Phase 116 DFT output (HDF5 format). Provide adapters for different DFT implementations (plane-wave, LCAO) to the unified wavefunction interface needed by spectroscopy modules
  • 124.11.1.2 Implement GW quasiparticle energy interface: read GW self-energy and quasiparticle corrections from Phase 118 output. Apply scissor operator or full quasiparticle corrections to band energies used in spectroscopy calculations
  • 124.11.1.3 Implement BSE exciton interface: read BSE exciton wavefunctions and eigenvalues for optical and core-level excitations. Enable reuse of BSE data across optical (Phase 123), EELS, and XAS modules

124.11.2 Phase 119 Integration (Lattice Dynamics)#

  • 124.11.2.1 Implement phonon mode interface: read phonon frequencies and eigenvectors at Gamma from Phase 119 dynamical matrix. Provide to IR intensity, Raman tensor, vibrational EELS, and Mossbauer Lamb factor calculations
  • 124.11.2.2 Implement Born effective charge interface: read Z* tensors from Phase 119 DFPT output for IR intensity computation. Validate acoustic sum rule: sum_kappa Z*_kappa = 0
  • 124.11.2.3 Implement phonon DOS interface: read total and atom-projected phonon DOS for Debye-Waller factor (EXAFS), mean-square displacement (Mossbauer), and thermal broadening computations

124.11.3 Phase 122 Integration (Defect Physics)#

  • 124.11.3.1 Implement defect state interface: read defect formation energies, charge transition levels, and relaxed geometries from Phase 122. Feed into PL (CC diagram), EPR (spin density at defect), and local spectroscopy (site-specific XAS/EELS for defect sites in supercell)
  • 124.11.3.2 Implement defect-specific spectroscopy workflow: given a defect supercell from Phase 122, automatically set up XAS/XANES calculation at the defect site, compute local DOS, predict core-level shifts, and compare with surrounding bulk-like sites

124.11.4 Validation Suites#

  • 124.11.4.1 Implement XAS validation: compute Cu K-edge XANES for bulk Cu (FCC) using FMS and compare with experimental spectrum — verify edge position within 1 eV, major features (white line, post-edge oscillations) reproduced. Validate EXAFS chi(k) against known Cu-Cu distance (2.556 A) and coordination number (12)
  • 124.11.4.2 Implement XPS validation: compute C 1s binding energies for diamond, graphite, C60, and CO2, verify chemical shifts match experimental ordering and magnitude within 0.5 eV
  • 124.11.4.3 Implement Raman validation: compute Raman spectrum of silicon (single peak at 520 cm^-1, T2g symmetry), diamond (1332 cm^-1, T2g), and alpha-quartz (multiple modes at 128, 206, 265, 355, 394, 401, 464, 697, 795, 808, 1072, 1085 cm^-1). Verify frequencies within 5 cm^-1 of experiment
  • 124.11.4.4 Implement NMR validation: compute 29Si chemical shifts for quartz, cristobalite, and coesite; verify ordering and magnitude against experimental solid-state NMR data within 3 ppm. Compute 27Al EFG for corundum and verify C_Q within 0.5 MHz of experiment
  • 124.11.4.5 Implement Mossbauer validation: compute isomer shift and quadrupole splitting for alpha-Fe (delta=0, Delta_E_Q=0), FeO (wustite, delta~0.93, Delta_E_Q~0.8), alpha-Fe2O3 (hematite, delta~0.37, Delta_E_Q~-0.20). Verify within 0.1 mm/s of experiment
  • 124.11.4.6 Implement EELS validation: compute low-loss EELS for bulk Al (free-electron plasmon at ~15 eV), Si (plasmon at ~16.7 eV), and diamond (band gap at 5.5 eV visible in loss function). Verify plasmon energies within 0.5 eV of experiment
  • 124.11.4.7 Implement EPR validation: compute g-tensor for Cu(II) in tetragonal coordination (g_parallel ~ 2.2-2.4, g_perp ~ 2.05-2.10) and verify against experimental EPR data. Compute hyperfine coupling A(63Cu) and verify order of magnitude
  • 124.11.4.8 Implement end-to-end spectroscopy pipeline test: starting from a CIF file for SrTiO3 (perovskite), run complete workflow: DFT ground state -> phonons -> XAS (Ti K-edge) -> EELS (Ti L-edge) -> Raman (zone-center modes) -> NMR (47/49Ti) -> verify all spectra are internally consistent and match known experimental data

<!-- prettier-ignore-end -->