Professor, Groupleader in Chemistry and Physics of Nanostructures, Dept. of Condensed Matter and Interfaces, Debye Institute for NanoMaterials Science, Universiteit Utrecht
Breaking the Wall of Seeing Atoms and Molecules. How Atomic Microscopy and Spectroscopy May Help our Understanding of Materials
In November 1989 I was in Utrecht.
Good afternoon ladies and gentlemen. There is one big dream in nanoscience: to relate the atomic structure of a system (this can be a molecule, quantum dot or nanowire, …) to its function. The function can be mechanical, chemical, or opto-electrical. For this purpose scientists have tried to image atoms in different situations. Here you see a picture acquired by Atomic Force Microscopy in our laboratory of a cupper surface. We see an ordered array of balls: that must be the copper atoms! Finally, 2500 years after the atom philosophy of Democritus, we are able to see atoms on the surface of copper. But, is this all? Should we content ourselves just to see the atoms as an ordered array of balls? I don’t think so. We should interrogate atoms in many different ways, which means exciting them in different ways and reading their response (measured signal), and learn as much as possible.
To explain my point, I show you this picture of a girl. The only thing we know is what we see: a girl. But we don’t have any other information: we don’t know who she is, what she is precisely doing, and what her interaction with the environment is. So, we need more information (that is more signals or responses to questions) to learn more about this girl. In fact, we have a similar situation with atoms and molecules. Besides the AFM microscopic picture of the balls, we need more ways “to speak” to individual atoms or molecules. That is precisely what I want to show you today.
We will talk about two types of “scientific seeing”: First, seeing in flat land, i.e. seeing atoms or molecules attached to a surface in two dimensions using scanning probe methods. Second, we will discuss how we see atoms in three dimensions, when they are organized in a crystal.
First: Scientific seeing in flatland. First, I will explain the method called scanning tunnelling microscopy. A scanning tunnelling microscope is, in fact, a very simple instrument. It is based on a needle with an atom end, placed very close to the surface investigated (distance is a few atomic radius). There is a voltage difference (bias) between the needle and the surface. That is the excitation. The measured response of the atoms (the signal) is the current between the needle (tip) and the surface. This current measures the number of electrons per unit time that tunnel through the space between the tip and the surface, called the tunnel barrier. In the picture, we indicate the tunnelling process with the green arrow. Electron tunnelling through a barrier depends sensitively on the width of the barrier. This means that the current basically flows between the last atom of the tip and the first atom of the surface underneath it. That is why, in principle, you could get something like atomic resolution with scanning tunnelling microscopy.
Lets look to the surface of the technologically important semiconductor, Indium Phosphide (InP) with scanning tunnelling microscopy. This picture presents a result of another research group. I have chosen it because it is very instructive on what we see with scanning tunnelling microscopy. We start with a voltage difference of - 2.7 V between the tip and the sample. What we see is a spatial picture of the tunnelling current as a function of the tip position over the InP surface. We see an ordered array of “blobs”, different from the Atomic Force Microscopy picture of the balls shown before. If we change the bias to positive values, we suddenly see a totally different pattern of “blobs” and the blobs also have a different shape. If we combine these two pictures, we get an ordered array of two different shapes of blobs, but it seems we don’t see the atoms at all!
So, we will try to understand what we are really seeing. We have to do with the signal, which is the tunnel current, and it goes from tip to the surface, or opposite. When the potential is negative, the electrons tunnel from the InP surface to the first atom of the tip: from a filled electron orbital of the InP to an empty orbital of the tip. Hence, what we image are the filled orbitals of the semiconductor surface. They are belonging to phosphor atoms. If we change the potential from negative to positive then also the tunnelling current is changed from direction. Now, electrons tunnel from the tip to the InP surface, into the empty orbitals. These empty orbitals are belonging to the Indium atoms. What we see, in fact, are not atoms, but we see an order array of Phosphor and Indium orbitals.
Second, we should explain scanning Atomic Force Microscopy. What do we see and why do we get a picture of atoms? I show again the picture of the copper surface, with the ordered array of atomic balls. Now we use the tip, not for measuring the tunnelling current, but we measure the force between the last atom on the tip and the surface. So, the signal we measure is now very different from the first case. To make the tip extra sharp sharp, we have attached a carbon monoxide molecule to the last atom of the copper tip, because this constitutes a much sharper and stiffer probe. But, the question of why we can see the copper atoms is not yet answered.
First I show you another example. This is an example of the IBM group in Zurich where scanning tunnelling probes were invented and awarded with the Nobel Prize. My colleague Dr. P. Liljeroth has tried to image a flat organic molecule, named pentacene, with Atomic Force Microscopy. We see a remarkable nice picture of the electronic skeleton of the pentacene molecule. The image is obtained again with a tip modified with a CO molecule.
Why can we get this type of atomic resolution? Well, the secret is more or less explained in this green panel there. It shows the images of the pentacene molecule for different distances between the CO molecule on the tip and the pentacene molecule on the surface. If the pentacene-CO tip distance is relatively large (2.6 Angstrom, a few atomic distances) the pentacene-CO tip force is attractive and consists of the Van der Waals force between the pentacene and the CO molecule. Because this force is not so strongly distance dependent, we are not able to see the atomic details of the pentacene molecule. But, if one positions the CO tip closer to the pentacene molecule, one begins to see atomic details in the molecule. At this shorter distance, the repulsion between the electrons of the CO and pentacene molecule becomes important. This repulsion force, also called chemical force, is very strongly distance dependent, and leads to an image of the electronic backbone of the pentacene molecule. We conclude that one needs a considerable strong repulsive force to obtain atomic resolution with Atomic Force Microscopy.
What happens to the CO tip at short distance and strong repulsion? That is we tried answer by experiments in our group. We have pushed the CO molecule of the tip as close as possible to a second CO molecule that stands upright on a copper surface. When the repulsive forces become too strong, the CO molecules start to buckle, and the bonding of the CO to the tip and surface becomes disturbed. This buckling indicates the ultimate resolution limit.
Second: Scientific seeing in three dimensions. Crystals consist of an ordered array of atoms in three dimensions. We have just learned that with scanning probe methods we could image the surface layer of atoms. But: can we image the atoms inside a crystal?
First I explain how Transmission Electron Microscopy works. We measure how an electron beam is transmitted through a crystal, and is measured underneath the crystal as a projection. One of these two-dimensional projections is presented here: it in fact presents the vertical voids between the atomic columns! Hence, the order that we see presents the lattice order. But, it is not more than a single projection in one direction. What do we have to do to break this wall of looking inside? Well, we have to make projections in many different directions. There is a mathematical theorem that says that if you have enough projections, it is possible reconstruct the total structure in three dimensions. The method is called tomography. In collaboration with with the group of Gustave van Tendeloo in Antwerpen we performed electron tomography. The precise technique is called HAADF electron tomography.
This is a movie showing you what the reconstructed result is if you do this with a small crystal of PbSe. We see a three dimensional picture of all lead-atoms in the crystal. If I rotate this reconstruction, it seems that we can look inside the structure. This is indeed the case. Hence, one can also find sites where one atom is missing. Hence, it is possible to study all sorts of defects. That is precisely what we need in order to fulfil the dream in nanoscience I emphasized in the beginning of this lecture: relating the atomic structure to the physical properties and functioning of a nano-system.
So, to end my talk, I recall that I have tried to tell you what scientific seeing really means. We started with a picture of a girl, and emphasized that more information is required to make this picture meaningful. In the case of “scientific” seeing of atoms or molecules, this can be achieved by looking to electron tunnelling currents or to the repulsive force. In the case of atoms in a crystal the sum of all transmitted patterns of the electron beam enable ones to reconstruct the three-dimensional atomic structure of a crystal: we can look inside!