In the Dept. of Synchrotron Radiation Research (Sljus) in Lund we study the surfaces of materials, a subsection of solid state physics often called 'Surface Science'. Although some of our work is relevant to modern industrial technology, most of it is pure science which aims to answer very basic questions about the nature and behaviour of surfaces. Today's surface science encompasses a very broad range of activities, from problems in catalysis and corrosion in the world of physical chemistry to studies of the geometric and electronic structure of semiconductor surfaces that are vital for the growing fields of nanotechnology and quantum electronics. The range of experimental techniques we use is similarly broad, as is the mixture of backgrounds of the scientists we work and interact with. Our particular speciality is Scanning Tunnelling Microscopy (STM), which is especially good at revealing surface structure in addition to representing a practical application of some very interesting theoretical problems.
Despite the fact that the essentials of quantum mechanics were sorted out in the 1920s, there are still no really good, predictive models for what goes on at the surfaces of even very simple crystalline materials. The problem is that the surface breaks the symmetry of the crystal lattice, and all sorts of rearrangements of the atoms and their accompanying electrons can take place because the atoms on the surface only have half as many neighbours as those deep inside the crystal. Many real-world properties of the material can and do change because of such surface rearrangements.
The STM can take atomically resolved pictures of the electron clouds surrounding surface atoms. It can tell the difference between electrons with different energies, and map their positions independently of each other. Thus STM is a very powerful tool for investigating surfaces, particularly when the information it provides can be cross-referenced with that given by other techniques.
The first picture shows an STM scan of a silicon (Si) crystal, processed by the computer to look like a three dimensional surface. The crystal is made up of a lattice of atoms arranged as if they occupied the corners of an infinite array of cubes (there are some atoms inside each cube too, but the overall packing pattern is cubic). The surface shown has been cut so that it is at a slight angle to the regular array of atoms, and consists of a series of steps and terraces aligned with the high-symmetry directions within the crystal. Imagine making a large rectangular block out of building bricks or Lego and then taking a thin wedge or slice off one side without cutting through any of the individual bricks. Because the blocks only have vertical and horizontal sides you will end up with a surface whose average slope is the small angle you wanted but which is made up of a whole load of steps, each step being one brick in height.
This is what you see here: just as you cannot have half a brick, you cannot have half an atom either, so this surface is made up of a series of steps, all or which are one or more atomic layers high. Steps running in one direction (the green ones) tend to be many atomic layers high, whereas those running in the other direction tend to be only one or two layers thick. Pictures like this can be related to the energies of different step shapes and directions and can tell us a lot about how the Si crystal grows, as well as how it will react when heated or when etched by corrosive chemicals.
The second picture shows a close up of one of the single layer steps. Each bright bump in the pattern is one Si atom. The beauty of this microscope is not just that it can see the atoms and their regular arrangement on the surface, but that it can also see defects such as the few positions where atoms are missing. The only other microscope that can resolve atomic detail, the electron microscope, uses diffraction from rows of atoms to form its image and so cannot detect a single missing atom like this.
Since the electrical conductivity of semiconductors like Si is largely determined by the defects in the materials rather than the properties of the pure lattice, this sort of image is very useful to people like chip makers who want to know exactly what is going on in their devices. Modern chips are made up of elements as small as forty of these pictures side by side, so even single atom defects can be important.
The symmetry of a crystal changes according to how you look at it. If you look at the side of a cube you see a square, but if you rotate it so that you are looking down on one of the corners you will see a hexagonal shape that has triangular symmetry. Both of the previous pictures are from a silicon crystal cut across the corner of the crystal lattice, which is why the angles between the steps are 120° and why the atoms in the atomic scale image are arranged in triangles. However, in this third picture we are looking at a Si crystal which has been cut parallel to the face of the cube, and the symmetry is square rather than triangular.
On this surface the atoms bond in pairs called 'dimers'. The dimers form parallel rows alongside each other, and swap direction by 90° when crossing a step. Both the parts of the image are of the same area on the surface and you can see the change in the dimer row direction very clearly.
Some of the dimers are tilted, with one half of the pair higher than the other. This is a sign of strain in the surface, usually caused by a step edge or the presence of defects. Defects such as whole missing dimers are regularly seen on this surface, as well as defects in which only one half of the dimer is missing. The remaining single atom in the second type of defect shows up very brightly in the right hand image, showing that the electrons around it behave differently from those around the normal dimer atoms.
The two halves of the picture are of the same part of the surface but look different because they show the electron distributions at different energies. The picture on the left is at an energy where the electrons are trapped between the two atoms of the dimer pair, and all you see is a sort of lozenge or oval. The picture on the right is at an energy where the electrons can sit on one atom or the other, but not in the middle, and you can now see two features in each dimer, corresponding to the two atoms in the dimer pair. This voltage-dependent nature of the STM pictures is a very powerful tool for surface analysis since it allows us to map the spatial distribution of electrons with a given energy and thus study the influence of defects, adsorbed gases or other interruptions to the crystalline surface.
Maintained by: Sljus Webmaster
[ Lund University ]
[ Fysicum ]
[ Maxlab ]
[ Sljus ]
[ STM Group ]
961217