Glossary

Index of commmon terms used in this website

• TEAS
• EPITAXIAL GROWTH
• LEED ( Low Energy Electron Diffraction )
• MEED ( Medium Energy Electron Diffraction )
• AES ( Auger Electron Spectroscopy )
• MOKE ( Magneto-Optical Kerr Effect )

TEAS

The diffraction of thermal energy He atoms was demonstrated shortly after the first observations of electron diffraction. However, the development of He diffraction as a technique for surface characterization had to wait until the 1970's. The main impulse came from the appearance of supersonic sources, which improved the intensity and monochromaticity of the neutral atomic beams by orders of magnitude.

An important characteristic of the TEAS technique, making it specially adequate for surface structure and growth studies, is its nil penetration depth. This is due to the low energies of the incident atoms, typically in the range 20-100 meV. He atoms are the most frequently used, because they are chemically inert, do not form molecules (which complicates the scattering process) and have a small mass. This technique has been mainly applied in three areas of research:

• Inelastic scattering, to investigate surface dynamics (surface phonon dispersion curves);
• Surface diffraction, applied to the study of ordered structures (especially those formed by light adsorbed species)
• Elastic scattering, to characterize disorder and defects at surfaces.

The latter applications take full advantage from the extreme sensitivity of the thermal He atoms to the presence of any kind of adsorbates or defects. This sensitivity is quantified in terms of a cross section for diffuse scattering with typical values of ~100 Å for single adatoms. Such large cross sections are related to details of the atom-surface interaction.

As shown in the image, the He-surface interaction potential has a long-range attractive part for large separation, and a short-range, repulsive component at small distance. The former is a dipole-dipole interaction due to quantum fluctuations in the charge density distribution, while the latter is caused by the overlap of the electron clouds of the He atoms with the outer electrons of the surface. The turning point of the He atoms is located about 3-4 Å from the surface atomic cores3.

The presence of an adsorbate or a defect strongly modifies the interaction potential. First, the electron charge distribution is altered, leading to changes in the repulsive component of the potential. But also the attractive part is modified, due to the He-adparticle dispersion forces. These forces are long-ranged, and therefore dominate the contribution to diffuse scattering, especially at very low defect concentrations, being thus the origin of the large cross sections.

All these characteristics make TEAS a very adequate technique for the study of atomic phenomena at surfaces, such as diffusion and growth. The kinematic nature of the atom-surface scattering process also facilitates the quantitative analysis of the data4.

EPITAXIAL GROWTH

When deposited onto a solid surface, except under conditions of extremely high deposition rates and/or very low temperatures, which may strongly limit diffusion, the incoming adatoms accommodate at the high symmetry positions dictated by the atomic structure of the substrate. This process is called epitaxial growth, because the overlayer reproduces, at least to some extent, the underlying surface ordering.

Taking advantage of this fact, it is possible not only to use the substrate as a support for the overlayer, but rather as a template to produce materials with metastable crystalline structures and exotic properties, not found in bulk phases.

In a first-order approximation, and assuming thermodynamic equilibrium, the growth morphology would be determined by the balance of the different surface free energies involved

Δγ_total = γ_a + γ_i - γ_s

where the subindices a, i, s stand for adsorbate, interface and substrate, respectively. Hence:

• Δγ_total ≤ 0: When the energetic cost of forming a substrate-adsorbate and an adsorbate-vacuum interface is smaller than the free energy of the substrate surface, covering this latter is favorable, and the deposit grows in a layer-by-layer fashion (FM or Frank-van der Merwe growth mode). In heteroepitaxial systems (those in which the adsorbate and the substrate are different materials) with large lattice mismatch, one must also consider the elastic strain energy accumulated with increasing thickness; this term is responsible for the break-up of the film into 3-dimensional islands after the growth of a few atomic layers (SK or Stranski-Krastanov mode).
• Δγ_total > 0: If the balance of energies for the growing film is positive, the system tries to maintain the substrate surface uncovered for as long as possible, while minimizing the exposed area of the deposit; this results in the formation of 3-dimensional islands (VW or Volmer-Weber mode).

Nevertheless, these thermodynamic criteria are frequently non-valid, first because growth from the vapor phase takes place at far-from-equilibrium conditions, and second because surface free energies are macroscopic magnitudes, not well defined for atomic-scale objects. Kinetic effects are extremely important as they allow us to obtain metastable structures. For these reasons, studies of growth in real time are crucial to fully appreciate the relevance of the different processes.

Diffraction techniques such as TEAS are specially adequate for this purpose, since they can provide with average views of the surface morphology and its time evolution. Fig. 2 shows schematically the typical behavior of a diffraction (surface reflectivity) experiment during different types of growth.

LEED ( Low Energy Electron Diffraction )

Due to the oscillatory nature of their wavefunctions, when a beam of electrons impinges on a surface with a periodic two-dimensional structure, a diffraction pattern appears in the same way as when electromagnetic radiation is used1.

This effect was first observed in 1927 by Davisson and Germer. In the energy range 10-1000 eV, the electrons interact strongly with the substrate atoms and lose energy rapidly after penetrating into the crystal. Therefore, filtering out all scattered electrons but those with kinetic energies very close to that of the primary beam ("elastic") yields information only on the uppermost surface atomic layers (1 to 10); the electrons mean free path can be estimated from a universal curve.

At the same time, such a strong interaction with the crystal atoms complicates the data analysis, due to the need to take into account dynamical and multiple scattering effects.

A typical LEED experiment:

• An electron gun (integrated within the whole LEED optics system) produces a monochromatic (?E ~ 0.1 eV) beam in the energy range between 10 and 1000 eV; this beam impinges onto the sample surface, usually at normal incidence.
• In order to remove all diffracted electrons that have lost energy through inelastic scattering events with the sample atoms, a retarding field with spherical symmetry is created by applying voltages slightly lower than the primary beam energy to the grids situated in front of the screen.
• The electrons that overcome this retarding field are accelerated toward the phosphorescent screen, where they produce bright spots whose intensity is proportional to the number of electrons in the corresponding beams.

The LEED pattern is an image of the reciprocal lattice. Intensity maxima appear at those points in reciprocal space that fulfil the two-dimensional Laue conditions. These can be illustrated in a graphical way by means of the Ewald construction: the sphere radius represents the wave vector of the incident electron beam, and diffracted beams appear wherever a reciprocal lattice rod intersects the Ewald sphere. The diffracted pattern thus reflects the symmetry of the surface unit cell, and the separation between the beams is inversely proportional to the interatomic distance.

Despite their surface sensitivity, the electrons used in LEED experiments penetrate a few atomic layers into the substrate, and therefore they can provide information on the subsurface structure. In practice, this is achieved by varying the primary beam energy (which, in turn, modifies the electron wavelength and their mean free path inside the solid) and measuring the changes in the diffracted beam intensity.

Experimental setup for LEED experiments: under computer control, the primary beam energy can be varied, while the diffracted patterns diplayed in the rear-view optics are acquired by the video camera and digitized at the framegrabber for processing.

MEED ( Medium Energy Electron Diffraction )

The MEED technique is very similar to the better-known RHEED (Reflection High-Energy Electron Diffraction), except for the lower energy of the incident electron beam, typically in the range 2-5 keV. Its usage is also similar: due to the higher energy of the electrons, grazing incidence is normally used to minimize their penetration into the sample and enhance the probe's surface sensitivity. The special geometrical arrangement makes the MEED or RHEED patterns look very different from LEED ones. Again in this case, the Ewald construction can be used to illustrate the scattering conditions and the formation of the diffraction pattern.

From the technical point of view, there are also some differences to point out with respect to the typical LEED setup1. With MEED or RHEED it is in general not necessary to bias the screen to accelerate the diffracted electrons, since their energy may be already high enough to produce fluorescence. Also, the intensity of the diffracted beams is much higher than the background, and therefore it is not necessary to filter out the inelastic electrons by using a suppressor voltage.

In our system, we use the integral electron gun of our CMA to produce a primary beam of 3 keV focused at the sample; the diffracted beams are then observed at the LEED screen, which is situated opposite to the CMA. The measurements can thus be performed using the same setup used for LEED.

AES ( Auger Electron Spectroscopy )

This is one of the earliest experimental techniques in surface physics, and it is nowadays routinely employed for chemical characterization. It owes its name to Pierre Auger, who first identified in 1925 the process of electron emission bearing his name today. In it, a deep electron is extracted by irradiation with either photons or electrons; the hole left by this electron is filled by another one falling from an upper shell, and the energy radiated in this recombination is absorbed by a third electron from the upper shells, which thus becomes able to leave the atom. This is a non-radiative process; the energy of the emitted electron depends only on the energies of the three levels involved, and it allows therefore for a unique determination of the emitting atom.

The typical kinetic energies of the Auger electrons lie between 20 and 1500 eV; for those values, the mean free paths of electrons are of the order of a few monolayers, and therefore the analysis of an Auger spectrum gives information on the chemical composition of the uppermost atomic layers of the material under study.

In our system, we use a CMA ("Cylindrical Mirror Analyzer") detector for the AES experiments. The excitation is achieved by an electron beam (~1 µA) produced by a 3 keV electron gun integrated within the spectrometer. This is a single-pass type; it uses a modulated voltage ramp to discriminate the energy of the electrons emitted from the sample; the resulting signal is enhanced by a channeltron and fed to a Lock-in amplifier. The data plotted in spectra corresponds to the first derivative (dN/dE) of the electron intensity with respect to the energy. The detection limit is typically 10-2 ML.

MOKE ( Magneto-Optical Kerr Effect )

During the 19th century, Michael Faraday and John Kerr observed that the polarization of a light beam was modified after being scattered off a magnetized material. For this reason, this phenomenon is nowadays called Faraday effect when it takes place in transmission, and Kerr effect in reflection. At the microscopic level, it is related to the spin-orbit coupling; its origin lies in the interaction between the electric field of the light and the magnetic moments in the solid.

While the magnitude of these effects in non-magnetic materials is too small for an efficient detection, in a ferromagnet such as Fe it can amount up to 3.8×105 deg./cm. Figures like this make them interesting candidates for magnetic data storage and readout. Additionally, combining the above mentioned values for the Faraday and Kerr rotations with the typical resolution of polarizers (10-2 - 10-3 deg.) allows for the detection of magnetization at the monolayer level. MOKE was first employed as an experimental technique for the characterization of the magnetic properties of surfaces and overlayers in 1986. The fact that most of the necessary equipment can be mounted outside the Ultra-High Vacuum chamber makes this technique particularly convenient for in-situ studies of magnetism in epitaxial systems.

The experiment starts with zero Kerr signal (obtained by crossing the polarizer and the analyzer) on the demagnetized specimen. Then, the intensity transmitted through the analyzer is monitored as an external magnetic field is swept upon the sample; any nonzero signal measured results from the rotation of the light polarization caused by the changes in the sample magnetization.

Different experimental modes are defined, depending on the relative orientation of the applied magnetic field, the scattering plane of the light beam and the surface magnetization. In the polar geometry, the experiment is sensitive to the perpendicular component of the sample magnetization, while in the longitudinal and transversal modes, the in-plane magnetization is detected.