Skip to main content Skip to navigation


Techniques for surfaces must be both surface specific and surface sensitive, thus techniques like neutron diffraction, in which the surface signal is dwarfed by the signal from the bulk, are unsuitable. This has led to an explosion in techniques to look at surfaces, many painstakingly designed in order to produce the most amusing acronym (for example SHMOKE and PhD which stand for Second Harmonic Magneto-Optic Kerr Effect and Photoemission Diffraction respectively).

In this project, the techniques of most interest are;

  • Low Energy Electron Diffraction (LEED)
  • X-ray Photoemission Spectroscopy (XPS)
  • Atomic Force Microscope (AFM)

Low Energy Electron Diffraction (LEED)

In LEED structural information is gained by using the sample as a diffraction grating for electrons of d’Broglie wavelength comparable to atomic separation. A schematic of a LEED rig is shown in figure 6. Surface specificity is achieved using low energy electrons which have a small inelastic mean free path and are unable to penetrate deep into the material (Seah and Dench 1976). For an ideal bulk terminated surface, the diffraction pattern seen is the reciprocal space equivalent of the surface 2D net. If a surface reconstructs (whether due to relaxation or an ad-atom over layer) the resulting LEED pattern is a combination of both the reciprocal substrate 2D mesh and the selvedge mesh and so more spots are seen. In theory it should be possible to use the resulting pattern to discern the relative lattice parameters between the selvedge and substrate mesh, but in practice other effects complicate matters. These are extensively discussed in (Woodruff and Delchar, 31-48). Figure 7 shows a LEED patterns for Si(001)-Sb at different temperatures. The right hand image is of most interest. At a glance, it would appear to have 4-fold symmetry.

This peculiarity is in fact the result of scattering from multiple domains. In a single domain, the unit mesh is a 2x1 reconstruction (2 fold symmetric), however, other domains may have a 1x2 reconstruction, which is equivalent to the 2x1 reconstruction rotated by 90°. Both are physically identical except for this rotation and so are equally likely to form. The resulting diffraction pattern is a combination of the 2x1 and 1x2 patterns.



Figure 5 (left)– LEED rig. The grid assembly acts as a high pass filter, letting through elastically scattered electrons as opposed to secondary and in-elastically scattered electrons whilst providing a field free region around the sample.

Figure 6 (below)– LEED pattern of Si(001)-Sb demonstrating a 1 x 1 reconstruction at T=450 °C (left) and a 2 x 1, two domain reconstruction at T=775 °C (right). (Mitsui, Hongo and Urano 2001)

LEED pattern


Reflection High Energy Electron Diffraction (RHEED)

In RHEED the large forward scattering cross section of high energy electrons (typically 12.5 keV) is taken advantage of by the geometry of the RHEED system (see figure 8). Forward elastically scattered electrons strike a phosphorus screen, while a grid assembly is used to cut out secondary and inelastically scattered electrons. Surface specificity is achieved by impinging the electrons at grazing angles to the surface along a certain crystallographic direction. The resulting RHEED diffraction pattern consists of a combination of diffraction streaks resulting from the substrate, integer order streaks (IOP), and the selvedge, fractional order streaks (FOP) (see figure 9). In order to sample different crystallographic angles, the sample must be rotated about its normal axis. The resulting data from the different Azimuth angles is used to determine the substrate lattice parameters and surface reconstruction [33]. Furthermore, RHEED provides a qualitative measure of surface quality. A well ordered smooth surface is characterised by sharp streaks, where as islands of material produce transmission diffraction spots on the pattern.


XPS is a non-destructive technique used to determine the elemental composition of a surface. Soft monochromatic x-rays of known energy... impinge on a surface and are absorbed by atoms, exciting them. By emission of a core electron, binding energy..., the atoms return to their ground state (Einstein 1905). The kinetic energy of the emitted electron is given by ....

Using a suitable analyser, the kinetic energy of the emitted electron can be found, enabling the determination of the binding energy for the electron. The core energy levels in a free atom are element specific, so it is possible to determine the element and shell identity from the binding energy.

There are numerous methods to determine the emitted electron energy (Prutton 1994, p 17). For this project a Concentric Hemispherical Analyser (CHA) was used. This technique of analysis is similar in many ways to mass spectrometry. The emitted electrons are focused by an electrostatic lens into a variable electric field provided by two concentric hemispheres (figure 7). Only electrons with a range of kinetic energy..., will make it through to the detector. Electrons with too high/low kinetic energy will crash into the sides of the hemispheres. The electric field is altered by varying the potential across the hemispheres, and can be scanned through a suitable range, resulting in a plot of intensity versus binding energy. Comparison of the location of peaks to binding energy reference values enables the identification of specific elements in the surface as well as molecules.


Figure 7 – XPS CHA schematic

Molecular determination from XPS

The molecular environment of a specific core electron changes its binding energy by an amount with respect to the binding energy of the electron in an atom [36]. Thus the final peak in an XPS spectrum is the sum of closely spaced spectral features from different molecules, which sum together to form the final XPS peak. For example the Ga(3p) peak is in fact a superposition of the Ga(3p-GaAs), the Ga(3p–Ga02) and others [37]. Information on the surface concentration of a particular element/molecule is encoded in the normalised peak height, given by....


AFM is used to map the surface topography of the sample. In constant force contact mode AFM [38], the repulsive force between a tip and surface is measured by the deflection of laser light off a cantilever attached to the tip. A feedback mechanism is employed in order to keep the force between tip and sample constant, by adjusting the height of the tip assembly. As the tip rasters across the surface, height deviation of the tip assembly is recorded and used to produce a plot of height as a function of position on sample. The use of piezoelectric scanners allows for sub nanometre adjustments in the height. Numerous other modes of operation exist, and are detailed in reference 39.