SPM-TH

Scanning Probe Microscopy Theory & Nanomechanics Group
Forces and Transport in Nanostructures

Research Highlights

Chemical identification of individual surface atoms by atomic force microscopy

image

Fig. 1 Schematic illustration of an atomic force microscope operated in a dynamic mode, and of the onset of the chemical bonding between the outermost tip a\ tom and a surface atom (highlighted by the green stick) that gives rise to the atomic contrast.





image

Fig. 2 Sets of tip-surface short-range chemical bonding force obtained over structurally equivalent Sn and Si atoms using identical acquisition and analysis\ protocols, but two different AFM tips. Although these set of forces are totally different, the relative interaction ration of the minimum force values for \ the two cures within a set remains nearly constant, with a value of 77%.





image
image
image


Fig. 3 (a), Topographic image of a surface alloy composed by Si, Sn and Pb atoms blended in equal proportions; two of the atomic species appear indistinguis\ hable as bright protrusions. (b), Distribution of the tip-surface minimum force values measured over the atoms in (a). By using the previously tabulated rel\ ative interaction ratio between pairs of atomic species (Sn and Si, and Pb and Si), each of the three groups of forces can be attributed to interactions mea\ sured over Sn, Pb and Si atoms. (c), Local chemical composition of the image in (a): blue, green, and red atoms correspond to Sn, Pb and Si, respectively.

The problem


The chemical identification of single atoms and molecules at surfaces has been pursued from the invention of both the scanning tunnelling microscope (STM) and the Atomic Force Microscope (AFM) The intrinsic detection nature of the STM and the AFM has hindered, until now, most of these efforts, and single atom chemical identification still remains a challenge.


In this work, however, we have operated the AFM in a dynamic mode, in a technique known as dynamic force microscopy (DFM). Under this scheme, the tip at the end of the cantilever is oscillated at a given amplitude and frequency (Fig.1), and these two magnitudes do change under the presence of a tip-surface interaction force. In our case, we keep the oscillation amplitude constant and recorder the variations on the oscillation frequency with changes in the tip-surface interaction force.


DFM is a very sensitive technique. In the most refined experimental set ups (as in our case), DFM allows one to detect and quantify the short-range chemical interaction force between two atoms. If the oscillating tip is driven close enough to the surface, so that the apex of the AFM tip gets closer that 5 angstroms (an angstroms is a ten millionth of a millimetre) during the turning point of the oscillation (Fig. 1), the onset of the chemical bonding between the outermost atom of the AFM tip and the individual atoms of the surface takes place (highlighted by the green stick in Fig. 1). This is, indeed, the mechanism behind the capability of DFM to truly image the atoms at insulator, semiconductor, and metal surfaces.


On the quest for single-atom chemical identification, DFM may have an advantage since, as it has been stated above, the imaging mechanism is based on detecting the short-range forces associated with the onset of the chemical bonding between the outermost atom of the AFM tip and the atoms at the surface.


Forces associated with the chemical bonding between two atoms are related to the nature of the atomic species involved. Thus, the short-range chemical forces we are measuring over the different surface atoms when exploring a heterogeneous surface with DFM should contain information about these surface atoms chemical nature. However, to extract this information is not trivial at all since, as we demonstrate in this paper, these short-range chemical forces present a strong variability upon the tip used to probe the surface, that is, for different AFM tip terminations we obtain unlike short-range chemical forces.


This variability is illustrated in Fig. 2, where the short-range chemical bonding force between the outermost atom of the AFM tip and two different atomic species at a surface, namely tin (Sn) and silicon (Si), is shown. The typical behaviour of these short-range chemical forces when approaching the tip towards the surface is a curve with an initial reduction of the force values from zero down to a minimum, from which the forces start increasing towards positive values; here, negative forces mean an attractive interaction between the atoms. As it can be seen, the set of force curves in Fig.2a is completely different from the set shown in Fig. 2b, even when they were very precise measured over a Sn atom and a Si atom with exactly the same acquisition an analysis protocol. The only difference between the two sets of curves depicted in Fig. 2a and 2b, respectively, is that they where measured using two different (unknown) terminations for the AFM tip.


In this publication, we also demonstrate, by means of state-of-the-art quantum mechanical simulations, that this variability in the measured short-range chemical forces is related to the different atomic structure and chemical termination of the AFM tip.


The findings


We have found a magnitude that remains nearly constant independently of the AFM tip termination we used. This magnitude is the relative interaction ratio of the minimum values of the short-range chemical forces measured over two different atomic species probed with the same tip (relative interaction ratio for short in the following).


This can be also seen in Fig.2. If we take the minimum force value for the curves measured over the Si atoms as 100%, the minimum force value for the curves measured over the Sn atoms will be in both cases close to the 77% (you can measure the length of the vertical lines beside the curves in each of the panels, and then divide the shorter by the longer).


We have corroborated this finding also for other atomic species like lead (Pb) and indium (In). We have measured the short-range chemical forces over Pb and Si (mixing the two atoms over the same surface) using different tips, and then we have quantified the relative interaction ration for these two species, resulting in a value of 59%. The same procedure, mixing In and Si on a surface in this case, leaded to a ratio of 72% for In and Si. In both cases (Pb-Si and In-Si), we found that the values of relative interaction ratio were almost independent of the AFM tip.


Furthermore, these values have been closely reproduced by means of state-of-the-art quantum mechanical simulations, with the calculation of the chemical bonding forces when probing Sn, Pb and In atoms of a surface using different atomic structures and chemical terminations (Sn, Si or Pb) to simulate the apex of the AFM tip.


The identification method


In this publication, we have demonstrated that the relative interaction ratio is connected to the relative strength these pairs of surface atoms have for making a chemical bond with the outermost atom of the AFM tip. This property makes it possible to use the relative interaction ratio as a fingerprint for the chemical identification of atoms at surfaces.


The identification method then consist on measuring the short-range chemical interaction force over each of the atoms in a surface area using the same AFM tip, and then compare the ratio of the minimum force values between pairs of atomic species with the previously tabulated relative interaction ratio for the expected atoms to be contained at the surface.


To demonstrate this method, we have used a surface alloy mixing Si, Sn and Pb in equal proportions, and identified each of the atoms in the imaged surface area (Fig. 3). When looking at the topography of this alloy (Fig. 3a) only one of the three species seems to present a different (diminished) contrast, while the other two cannot be differentiated. After systematically measuring the tip-surface, short-range chemical boning force over each of the atoms, we can see that the minimum force values registered over these atoms can be clearly classified into three groups, as it is shown in the histograms in Fig. 3b. When taking the previously tabulated values of the relative interaction ratio into account (77% for Sn and Si, and 59% for Pb and Si), these groups can be assigned to forces obtained over Sn, Pb, and Si atoms, and therefore each surface atom can be associated with the corresponding chemical element (Figs. 3c).


Relevance and possible applications


As mentioned above, this capability of identifying atoms at surfaces could multiply the already outstanding possibilities that DFM offers. This method might be of relevance in surface chemistry, material science, nanoscience and nanotechnology, and even in semiconductor technology; in particular, when combining this identification method with the ability of DFM for the manipulation of individual atoms at surfaces (a capability that we have recently also demonstrated).


In the field of surface chemistry and catalysis, the possibility of identifying and manipulating different species at surfaces will allow to produce and study controlled chemical reactions at atomic scale, identifying both reactants and products. In material science, this identification technique will allow, for example, to disclose chemical compositions of local surface areas and relate it to the local mechanical properties of the material at the nanometric scale.


In nanotechnology and nanoscience, the combination of atomic identification and atomic manipulation will enable to construct, atom-by-atom, nanostructures and nanodevices with specific properties and functionality, by mixing different atomic species while having an absolute control on the location and chemical nature of each atom. In semiconductor industry, the capabilities of DFM to identify and manipulate individual atoms at semiconductor surfaces will allow, for instance, the selective doping of semiconductors arranging dopants of different nature in particular arrays to enhance the performance of nanoscale transistors and, therefore, electronic devices.





Yoshiaki Sugimoto 1, Pablo Pou 2, Masayuki Abe 1, Pavel Jelinek 4, Ruben Perez 2, Seizo Morita 1 & Oscar Custance 1
Complex patterning by vertical interchange atom manipulation using atomic force microscopy
Nature 446 , 64-67 (2007).    (DOI link)    (pdf)


1 Graduate School of Engineering, Osaka University, 2-1 Yamada-Oka, 565-0871 Suita, Osaka, Japan.
2 Departamento de Fisica Teorica de la Materia Condensada, Universidad Autonoma de Madrid, 28049 Madrid, Spain.
3 PRESTO, Japan Science and Technology Agency, Saitama 332-0012, Japan..
4 Institute of Physics, Academy of Sciences of the Czech Republic, Cukrovarnicka 10, 1862 53, Prague, Czech Republic.