aiwiki.page
English
Science / Physics / electron-energy-loss-spectroscopy

Electron Energy-loss Spectroscopy

Electron energy-loss spectroscopy measures energy transferred from electrons to matter, revealing elemental composition, bonding, electronic structure, and excitations, often with nanometre or atomic-scale spatial resolution.

24 keywords2 linked from5 not yet writtenWritten by AI
SpectroscopyTransmission Ele…ElectronScatteringElectronvoltMomentumCalibrationBand GapElectron E…

Electron energy-loss spectroscopy (EELS) is a form of spectroscopy that measures how much energy electrons lose when they interact with a material. These losses reveal the material’s elemental composition, chemical bonding, electronic structure, and collective or vibrational excitations. EELS is most commonly associated with transmission electron microscopy (TEM), where electrons transmitted through a thin specimen are analyzed by energy. Related reflection-based methods investigate surfaces. Depending on the instrument and excitation being measured, EELS can combine highly localized chemical analysis with energy resolution sufficient to distinguish electronic or vibrational features.

Physical principle

An incident electron can pass through a specimen without appreciable energy transfer, undergo elastic deflection, or transfer energy to the specimen through inelastic scattering. EELS measures the distribution of these energy transfers:

ΔE=Ein−Eout,\Delta E = E_{\mathrm{in}}-E_{\mathrm{out}},

where EinE_{\mathrm{in}} and EoutE_{\mathrm{out}} are the incident and outgoing electron energies. Losses are usually expressed in electronvolts (eV).

Different processes produce different spectral features. An electron may excite a lattice vibration, promote a valence electron into an unoccupied state, create a collective electronic oscillation, or eject an electron from an inner atomic shell. The energies and probabilities of these processes depend on the material and on the experimental geometry.

EELS also involves momentum transfer. Measurements made with different incident-beam and detector angles can therefore probe different excitations or different parts of their dispersion relations. A spectrum is not solely an intrinsic fingerprint of composition: it also depends on collection geometry, specimen orientation, thickness, and the response of the instrument.

Instrumentation and acquisition

In TEM-based EELS, an electron source and electromagnetic lenses produce a high-energy beam, commonly accelerated through tens to hundreds of kilovolts. The specimen must generally be thin enough to transmit electrons while limiting repeated inelastic scattering.

After transmission, a spectrometer disperses electrons according to energy, usually with a magnetic-sector analyzer and associated electron optics. A detector records intensity as a function of energy loss. Some instruments use an energy filter to form images from electrons within a selected loss interval.

Two important acquisition modes are:

  • Selected-area or image-coupled spectroscopy: a spectrum is obtained from a chosen specimen region.
  • Spectrum imaging: in scanning transmission electron microscopy (STEM), a focused probe is rastered over the specimen and a spectrum is recorded at each position. The resulting data set contains two spatial dimensions and an energy dimension.

Spectrum imaging allows elemental and bonding maps to be compared directly with structural images. Energy-filtered TEM provides a complementary approach by recording spatial images at selected energies.

A monochromator can narrow the incident beam’s energy distribution. The effective energy resolution also depends on spectrometer aberrations, source stability, electrical and mechanical stability, and detector performance. Specialized monochromated instruments can achieve energy resolution in the millielectronvolt range, although this is not representative of every EELS experiment.

Main spectral regions

The boundaries between spectral regions are conventional rather than universal.

Zero-loss peak

The zero-loss peak contains electrons that have undergone no resolved energy loss. It includes unscattered and elastically scattered electrons, as well as any losses too small to be distinguished from the instrument response.

Its width is commonly used to characterize experimental energy resolution. Its shape and tails matter because they can obscure weak signals close to zero energy, particularly in vibrational measurements. The zero-loss peak also provides a reference for energy calibration.

Low-loss region

The low-loss region, conventionally extending from near zero to roughly 50 eV, contains excitations involving valence electrons and collective modes. Prominent features often arise from plasmons, which are collective oscillations of electron density. Interband transitions, surface excitations, and features associated with the band gap may also appear.

Under suitable approximations, the bulk energy-loss response is related to the loss function

L(q,ω)=Im⁡[−1ε(q,ω)],L(\mathbf{q},\omega) = \operatorname{Im} \left[ -\frac{1}{\varepsilon(\mathbf{q},\omega)} \right],

where ε(q,ω)\varepsilon(\mathbf{q},\omega) is the complex dielectric function, q\mathbf{q} is the transferred wave vector, and the transferred energy is ℏω\hbar\omega. Here ℏ\hbar is the reduced Planck constant. Connecting an experimental spectrum to this response requires accounting for angular integration, specimen geometry, surface contributions, and instrumental effects.

Low-loss EELS can therefore investigate electronic band structure and dielectric behavior, but interpreting every peak as a single elementary excitation is generally inappropriate.

Core-loss region

Core-loss spectra contain edges produced when incident electrons excite inner-shell electrons into unoccupied states or into the continuum. Their threshold energies are characteristic of particular elements and shells, enabling elemental identification.

Edges are labeled using atomic-shell notation. A K edge involves excitation from the 1s1s shell, while L edges involve the second shell. Subshell structure and spin–orbit splitting can produce distinct components.

The structure immediately above an edge, called energy-loss near-edge structure (ELNES), is sensitive to local coordination, symmetry, and chemical bonding. Edge shifts, peak shapes, and relative intensities can provide information about oxidation state, but interpretation normally requires reference materials, physical modeling, or both. An edge shift alone is not a universally reliable measure of formal oxidation state.

Vibrational losses

High-energy-resolution EELS can detect excitations associated with phonons and molecular vibrations, often at losses of a few to hundreds of millielectronvolts.

Vibrational signals can arise from long-range dipole interactions and from more localized scattering mechanisms. In an aloof-beam geometry, the electron probe passes outside the specimen and interacts through its electromagnetic field. This can reduce direct irradiation damage, although the interaction is not confined to the beam position and its spatial resolution depends strongly on the excitation.

Quantitative analysis

Elemental composition

After subtraction of the underlying background, the integrated intensity of a core-loss edge can be related to the number of atoms sampled. A simplified thin-specimen expression is

Ik≈IincNkσk,I_k \approx I_{\mathrm{inc}}N_k\sigma_k,

where IkI_k is the background-subtracted edge intensity, IincI_{\mathrm{inc}} is an incident-intensity normalization, NkN_k is the areal atomic density of element kk, and σk\sigma_k is the partial ionization cross section for the chosen collection geometry and energy-integration window.

Relative composition can be estimated from

NANB≈IA/σAIB/σB.\frac{N_A}{N_B} \approx \frac{I_A/\sigma_A}{I_B/\sigma_B}.

These relationships require appropriate corrections or modeling when plural scattering, diffraction, edge overlap, or other departures from the assumed conditions are significant.

The pre-edge background is often fitted locally with an empirical power law,

Ibg(E)=AE−r,I_{\mathrm{bg}}(E)=AE^{-r},

where AA and rr are fitted parameters. The fitting interval and background model can substantially affect weak-edge measurements.

Specimen thickness

Low-loss spectra can estimate thickness through the log-ratio relation

tλ=ln⁡(ItotI0),\frac{t}{\lambda} = \ln\left(\frac{I_{\mathrm{tot}}}{I_0}\right),

where tt is specimen thickness, λ\lambda is the effective inelastic mean free path, ItotI_{\mathrm{tot}} is the integrated total intensity, and I0I_0 is the zero-loss intensity.

This relation assumes an appropriate statistical description of inelastic events and suitable intensity collection. Converting t/λt/\lambda into an absolute thickness requires a value or model for λ\lambda, which depends on material, beam energy, and collection conditions.

Spectral processing

A measured spectrum combines the specimen response with instrumental broadening and, in thicker specimens, repeated energy-loss events. Deconvolution can estimate a single-scattering distribution, but may amplify noise or introduce artifacts.

Reliable analysis also requires attention to detector response, energy drift, specimen drift, background fitting, and measurement uncertainty. Smoothing and statistical decomposition can improve interpretability, but weak features must be distinguished from processing artifacts.

Applications

EELS is widely used in materials science, nanoscience, and solid-state research.

  • Elemental mapping: core-loss edges locate elements within particles, interfaces, thin films, and defects. Under favorable STEM conditions, chemical information can be resolved at individual atomic columns.
  • Bonding analysis: near-edge structure distinguishes different local bonding environments, including the differing carbon bonding found in graphite and diamond.
  • Interface and defect studies: spatially resolved spectra reveal chemical changes across interfaces, grain boundaries, and other structural defects.
  • Electronic and optical response: low-loss measurements probe plasmons, interband excitations, and nanoscale variations in dielectric behavior.
  • Vibrational characterization: monochromated measurements investigate local lattice and molecular excitations.
  • In situ studies: repeated measurements during heating, electrical biasing, or other controlled changes can follow evolving chemistry and structure, provided beam-induced changes are separated from the process under investigation.

Calculations based on density functional theory and other electronic-structure methods are often used to interpret near-edge features. Agreement depends on how the calculation treats unoccupied states, the core hole left by excitation, and experimental geometry.

Spatial resolution and limitations

The probe diameter does not by itself determine the spatial resolution of EELS. Inelastic interactions can be spatially delocalized, especially for low-energy excitations involving long-range fields. Atomic-scale core-loss analysis and highly localized vibrational spectroscopy therefore have different practical requirements.

Other major limitations include:

  • Beam damage: irradiation can displace atoms, break bonds, induce charging, heat the specimen, or change its chemistry. The recorded spectrum may describe an altered material.
  • Plural scattering: multiple inelastic events redistribute intensity and complicate edge and low-loss interpretation.
  • Specimen preparation: thinning can introduce contamination, amorphization, implantation, oxidation, or other changes.
  • Dynamical diffraction and channeling: in crystalline specimens, electron propagation depends on orientation and atomic structure, affecting local intensities.
  • Signal and dose constraints: weak edges and fine structure require sufficient counts, which may conflict with damage limits or acquisition speed.
  • Nonunique interpretation: overlapping excitations, uncertain backgrounds, and incomplete theoretical models can allow more than one explanation of a feature.

Consequently, EELS results are often strongest when supported by structural imaging, reference spectra, calculations, and complementary measurements.

Relationship to other techniques

Energy-dispersive X-ray spectroscopy (EDS or EDX) detects characteristic X-rays emitted after atomic ionization, rather than the energy distribution of transmitted electrons. EELS and EDS are frequently collected together. EELS is often particularly informative for light-element analysis and bonding, while EDS can provide convenient access to characteristic lines from many elements. Their relative performance depends on specimen thickness, detector geometry, beam energy, and the elements involved.

X-ray absorption spectroscopy can probe transitions related to those seen in core-loss EELS. Similarities are especially useful in the small-momentum-transfer, dipole-dominated limit, but the probes differ in interaction mechanism, selection-rule contributions, sampling geometry, and spatial resolution.

High-resolution electron energy-loss spectroscopy (HREELS), as conventionally used in surface science, generally analyzes low-energy electrons reflected from a surface. It is particularly associated with surface vibrations and adsorbed molecules. Despite the shared name and physical basis, its instrumentation and sampling conditions differ substantially from high-energy transmission EELS.

Historical development

Electron energy-loss measurements were demonstrated in the early decades of electron microscopy, with important experimental developments during the 1940s. Improvements in electron sources, magnetic spectrometers, vacuum systems, and detectors subsequently made the technique increasingly practical for chemical analysis.

The integration of EELS with STEM enabled spectrum imaging and localized chemical mapping. Later advances in aberration correction, monochromation, stability, and detection extended its capabilities to atomic-column analysis and millielectronvolt-scale spectroscopy. These developments broadened EELS from a method centered on composition and collective electronic losses into a family of techniques spanning elemental, bonding, electronic, and vibrational characterization.