Research activities:

Our research activities are currently targeted at four topics gathered around the nc-AFM technique in UHV.

 

·        Organic molecules on ionic crystals:

Owing to their large optical gap, bulk insulators are natural candidates for applications in optoelectronics involving molecular layers at the surface. Their weak surface reactivity allows for the electronic decoupling of the molecular layers from the substrate, while preserving their intrinsic functional properties, which is an important issue for functional nano-devices. Alkali halides single crystals are particularly interesting as they sometimes allow for the growth of well-extended molecular networks, as demonstrated in the case of perylene on KCl(001) (T. Dienel et al., Adv. Mat. 20, 959, 2008). For this system, it has been demonstrated that the molecular layer displays much sharper absorption spectra than in solution owing to its commensurate growth on the substrate and the subsequent minimization of the inhomogeneous broadening.

With these ideas in mind, we started investigating the molecule of benzene diboronic acid (BDBA) on KCl(001). In the volumic phase, BDBA molecules are known to form a lamellar structure, wherein the cohesion within each lamella is ensured by hydrogen bonds. When deposited on Ag(111) at room temperature, it has been demonstrated in the Nanostructuration group that the molecules polymerize to form a robust bi-dimensional covalent organic framework. When deposited on KCl(001) at room temperature, the molecules do not polymerize, but rather form a bi-dimensional supramolecular phase which is stabilized by hydrogen bonds. The phase consists of parallel rows of densely-packed molecules, 0.52 nm far apart from each other. The rows are separated by 1nm. The structural analysis shows that there is one molecule per unit cell that develops four H-bonds with the neighboring molecules. The growth of the supramolecular phase is only made possible by the conformational adaptation of the molecule that tilts its central benzene ring and hence, allows for the strengthening of the intermolecular H-bonds. This bi-dimensional phase is almost analogous to the one observed in individual lamellae of the volumic phase.

DFT calculations performed by V.Oison and M.Sassi from the theory group at the IM2NP have confirmed the supramolecluar structure and have computed a cohesion energy for it of about 0.95eV per unit cell, corresponding to almost 0.25eV/H-bond (5.2kcal/mole), i.e. intermediate-strength H-bonds.

 

Fig.2 a- (35x25)nm2 nc-AFM image of BDBA molecules on KCl(001). The molecular layer is visible on the left-hand of the figure and the atomic resolution on KCl is visible on the right-hand side. b- nc-AFM image of the supramolecular phase of BDBA showing dense rows of molecules. The unit cell is rectangular and consists of one molecule developing four H-bonds with the neighboring molecules. c- Structural model of the supramolecular phase showing the rectangular unit cell with the four H-bonds engaged by each molecule. It must be noticed that such a phase is allowed to the rotation of the benzene ring w.r.t. the plane of B(OH)2 groups.

 

·         Organic molecules on 6H-SiC(0001):

This project is funded by the “Agence Nationale pour la Recherche ” (MolSiC project, ANR-08-NANO-030-02) and started in January 2009. It is a fundamental research project, the aim of which is to explore the use of a large bandgap semiconductor, silicon carbide (SiC), for applications in molecular electronics. The suitability of this substrate for molecular electronics is that it allows for the interfacing between semiconductors technology and molecular electronics. But owing to its wide bandgap (up to ~ 3eV), SiC insures an adequate electronic decoupling between the molecules and the substrate as well. MolSiC sticks to three guide-lines: 1- synthesis of new and functionalized-on-purpose organic molecules. 2- structural characterization of the molecular nanostructures on the surface and electronic structure analysis. 3- combination between the latter experimental results, DFT calculations of the surface including the molecules and nc-AFM image calculations by means of recent numerical tools (see item “Modelization” below).

A large consortium of French research groups are involved in the project, among which the group Nanostructuration at the IM2NP. Its main task is to undertake a comprehensive study of the adsorption of the functionalized molecules at the surface of the 6H-SiC(0001) with the goal to trigger their ordered growth. We expect to benefit from the wealth of surface reconstructions of the 6H polytype of SiC, among which the 3x3 and the (√3x√3)R30°. First, following the bottom-up approach, the different adsorption stages will be investigated up to the monolayer regime in LEED, nc-AFM and inverse photoemission. The Nano-AFM group started the first nc-AFM; LEED and Auger investigations in September 2009 on the bare 6H-SiC(0001). In Fig.3 are given the first LEED patterns that have been obtained. nc-AFM characterizations of the 3x3 surface reconstruction are currently in progress.

 

Fig.3- LEED patterns of the reconstructions of the 6H-SiC(0001) surface recorded at 115 eV. From left to right: 3x3, (√3x√3)R30° and (6√3x6√3)R30°. For each pattern, the 1x1 unit cell of the surface is shown in red

 

·        Modelization:

1.      Analytical approach to the local contact potential difference (LCPD) probed by Kelvin Probe Force Microscopy on the atomic-scale

We have developed an analytical approach to the LCPD with the goal in mind to better understand the atomic-scale contrast observed while Kelvin Probe Force Microscopy measurements (KPFM) on KBr(001) (F.Bocquet et al., Phys.Rev.B 24, 1791 (2008)). The approach relies on the determination of the electrostatic force occurring between a metallic tip carrying a small metallic cluster in topmost position and the bulk ionic crystal below. Our approach, which is self-consistent and based on classical electrostatics, states that the ions at the surface undergo dynamic displacements (ionic polarization) under the influence of the inherent modulation of the bias voltage between the tip and the crystal that is required ny the KPFM technique. The force, with a short-range nature, is site-dependent and has the lateral periodicity of the Madelung surface potential, i.e. the one of the underlying ionic lattice. It does not only scale quadratically with the applied bias voltage, but also linearly. The analytical approach to the KPFM method, i.e. actually the analytical expression of the DC potential that nullifies this force, namely the LCPD, shows that the latter force is responsible for the atomic-scale contrast in KPFM but that any quantitative connection to the physical observables such as the Madelung surface potential or the local work function can be made owing to the convolution of the detected LCPD by the tip’s geometrical parameters.

 

Fig.4- Overview of the assumptions required to perform the analytical calculation of the electrostatic force in KPFM

Fig.5- Analytical expression for the electrostatic force which exhibits a short-range character as well as the lateral periodicity of the underlying ionic lattice. This force is responsible for the atomic-scale contrast reported in KPFM.

 

2.      Simulating the nc-AFM/KPFM setup.

Following the analytical developments, we have implemented the KPFM operational mode within the core of an accurate numerical implementation of the nc-AFM setup, the so-called nc-AFM simulator (L.Nony et al., Phys.Rev.B 74, 235439 (2006)). With the fully numerical nc-AFM/KPFM setup, it is possible to simulate spectroscopic curves as well as topographical and LCPD images. In collaboration with Prof. Adam Foster from Tampere University ( Finland ), we have used the nc-AFM/KPFM simulator in combination with atomistic force fields computed between a tip with a realistic geometry and the NaCl(001) surface. This large computational effort strengthen most of the analytical results and led to the conclusion that the simultaneous occurrence of the atomic-scale topographical and KPFM contrast is possible (no experimental artifact, no cross coupling between channels), but that the latter relies on the dynamic polarization of the ions at the tip-surface interface. This makes any quantitative measurement of the Madelung surface potential possible. However, the “chemical” periodicity remains preserved.

 

Fig.6- Sketch of the numerical nc-AFM/KPFM setup, based on the experimental implementation of the FM-KPFM setup

Fig.7- a- Sketch of the numerical tip-surface setup. We have set zm = 5 mm compared to z which scales in the sub-nm range. b- Sketch of the NaCl unit cell showing the 17×17 mesh used to calculate the (x,y,z,V) four-dimensional tip-surface atomistic force field. We have focused on four particular sites: anionic (A), cationic (C) and hollow (H1, H2) sites.

Fig.8- a- Topographical image computed with the nc-AFM/KPFM simulator. The vertical contrast is 38 pm. b- Simultaneously computed LCPD image. The contrast ranges from -2.24 to -1.69 V (0.56 V full scale). c- LCPD image computed at constant height, z = 0.45 nm. The contrast ranges from -2.24 to -1.38 V (0.86 V full scale), consistently with the expected range deduced from Fig.3b. d- Evolution of the magnitude of the LCPD contrast (dots) and of the average LCPD (squares) as a function of the distance.

 

·        Experimental developments: ultra-fast cantilevers for nc-AFM:

This project is funded by the “Agence Nationale pour la Recherche ” (NanoSens project, ANR-08-NANO-017) and started in January 2009. Its goal is to significantly improve the performance of non-contact atomic force microscopy (NC-AFM) by developing new force sensors (or cantilevers: CLs). In NC-AFM, the minimal detectable force, fixed by the thermal fluctuations of the CL, is given by:

For a given temperature kBT and bandwidth B, the sensitivity of the instrument can be improved by increasing the resonance frequency ω0 and/or the quality factor Q. Reducing the stiffness k is not possible below a certain limit, due to the occurrence of a jump-to-contact instability when the tip gets close to the surface.

1.      Increasing the resonance frequency of CLs while keeping k at values compatible with NC-AFM requires a decrease of the effective mass of the device, hence a size reduction. Here, we aim at developing CLs with submicronic dimensions (typically 1 x 0.4 x 0.1 μm3) from monocrystalline silicon carbide 3C-SiC films grown on Si. The excellent mechanical properties of SiC, when compared to Si, combined with the small size of these devices allow to reach resonance frequencies as high as 100 MHz, leading to a 100-fold increase relative to the silicon cantilevers in use nowadays.

2.      Increasing the quality factor will be done by optimizing the SiC material properties through a better understanding of the role of the bulk and surface defects in the dissipation of mechanical energy. We expect that these improvements will be beneficial well beyond NC-AFM applications, for SiC-based MEMS/NEMS and power microelectronics.

3.      The small size of these nano-CLs disqualifies the usual optical techniques for measuring their motion. The method to detect the CL displacement will be based on the strain-induced variations of the resistance of a thin metallic film deposited on the CL. This metallic piezoresistivity based approach is radically different from the semiconducting piezoresistivity based approach that has been used for about 10 years for AFM CLs. As demonstrated recently, it presents distinct advantages that will be fully exploited in the present project.

4.      The increase of the resonance frequency for NC-AFM so far was not only limited by the absence of appropriate fabrication facilities for the force sensors but also by the limited frequency range of commercial electronics (max. ~ 5 MHz). Extending the frequency range up to 100 MHz involves totally different technical solutions that will be developed in the project.

5.      Finally, the optimization of the instrument and its control system requires extensive simulations, which will be carried out by extending an existing "virtual NC-AFM" program to CLs with resonant frequency in the 100 MHz range.

 

In collaboration with H.Barthélémy’s group at the IM2NP (circuits design team), the Nano-AFM group is in charge to develop a high-speed PLL for being used with these integrated cantilevers.