Advanced Electrical and Mechanical Atomic Force Microscopy Characterization from Concept Scientific Instruments

Concept Scientific Instruments (represented by Mi-Net Technology Ltd in the UK & Ireland) is a scientific equipment manufacturer specializing in Atomic Force Microscopy. In this article by Louis Pacheco and Nicolas F. Martinez we discuss the use of modes for advanced electrical and mechanical AFM characterization.

Recent advances with brand-new materials such as graphene require new high resolution and sensitive characterization techniques. Concept Scientific Instruments has developed an alternative measurement mode called Soft Intermittent Contact mode (or Soft IC) that combines the advantages of contact mode and force spectroscopy without the inconveniences of friction forces or intrinsic slowness. Soft IC mode works as follows (Fig. 1(b)) : The tip is separated a safe distance (referred as « lift height ») from the surface and moved quickly to the next measurement point (determined by the number of pixels of the image), when it reaches the measurement point the tip is stopped and it performs a force spectroscopy curve. The maximum force applied is set by the user by a setpoint value in a similar manner as in contact mode. Another advantage of Soft IC mode compared to contact mode is the avoidance of instabilities due to changes in the adhesion force during the scan. By setting a lift height higher than the adhesion force, the tip can be totally disengaged from the surface. This also has the advantage that even softer cantilevers (typically they experience very high adhesion forces) can be used routinely with this mode. In addition, stiffness and adhesion can be obtained directly from every measured point. The stiffness (i.e., the ratio of the applied force and the deformation of the sample) can be used in combination with the Soft Meca software module to calculate the Young modulus.

An example of the advantages of Soft IC mode is illustrated in Figure 1(a). A PS/PMMA blend sample was first imaged in contact mode with a stiff cantilever (k = 37 N/m, ACT from AppNano). The surface shows two clear separated phases of PS islands (yellow domains) embedded in PMMA matrix (brown domains). Although the applied force (45 nN) is not high enough to produce a permanent deformation, both adhesion and friction forces produced during the scan drag parts of PMMA on both the PS islands and on the sides of the scanned area (white spots along blue-dashed square on Figure 1). The same force was applied using Soft IC mode (red square) while increasing the lift height up to 300 nm to completely release the tip from the surface. The image in Soft IC mode does not show permanent damage of the sample. The bottom part of the image corresponds with a measurement in resonant mode (green-dashed square) showing no permanent damage either. However, in resonant mode it is not straightforward to provide the value of the applied force. For the experimental conditions used, we estimate the applied average force of 38 nN.

Figure 1. (a) comparisons of Contact, Resonant and Soft IC modes on PS/PMMA sample. (b) features of Soft IC mode.

In addition, Soft IC can be combined with other measurement modes such as the “ResiScope” mode that allows characterization of electrical properties over 10 orders of magnitude (from 50 fA to 1 mA). In this way we can work in Soft ResiScope mode, therefore reducing the friction and wear on the conductive tip.  In Figure 2 (left side) is an example of a resistance measurements made on a metallic sample (blue areas- KΩ) with some corrosion on it (yellow and red areas- GΩ range) with an array of nanoindentations made on it. Due to the corrosion film, resistance is very high except for the areas surrounding the nanoindentations where the conductive AFM tip can make contact with the metallic surface. Figure 2 (right side) shows a cross-section along the labeled nanoindentations from 1-5. The topography cross-section shows some pile-up of material at the borders. The resistance cross-section shows that nanoindentations 1 and 5 have a larger conductive area (4 microns length), while 2-4 have smaller conductive areas (1.8 microns length) as they are located in the highly corroded areas. It also underlines the importance of corrosion on the final properties of materials, as one would expect the metal surface to have high conductivity. Due to the corrosion on it (which can naturally occur when exposed to oxygen or other corrosion agents) the conductivity decreases several orders of magnitude. The expected conductivity of the material is retrieved when measuring the exposed deeper layers with the indentation prints.

Figure 2: (Left) Atomic Force Microscopy 3D image with overlay of resistance measurement of a corroded gold sample. Red and yellow areas correspond to GΩ range while blue areas to KΩ range. (Right) topography and resistance cross-sections along the indicated nanoindentations.

Other electrical modes like High Definition-Kelvin Force Microscopy (HD-KFM) can be used with the Nano-Observer AFM for full electrical characterization. This technique allows simultaneous measurement of topography and surface potential at the nanoscale. Surface potential provides information of work function (an intrinsic property), trapped charges, dipoles,etc. HD-KFM excites the first two natural frequencies of the cantilever that corresponds to the first natural flexural eigenmodes as depicted in Figure 3. Thus, first feedback (mechanical excitation) is tuned to the first eigenmode frequency to measure topography (as in regular resonant mode operation), while a second feedback is tuned to the second eigenmode frequency that controls the Vac+Vdc bias and allows measurement of the surface potential (see Fig. 3).

Figure 3. Concept of multifrequency approach of HD-KFM. The first mode of the cantilever is excited mechanically (for topogaphy) while the second mode is excited electrically (for surface potential).

The advantages of tuning the VAC bias used for the electric feedback to the second eigenmode of the cantilever is that the signal is amplified by the Q factor of the second eigenmode. This effect provides the possibility to use smaller VAC values to obtain an oscillation amplitude with an acceptable signal to noise ratio as compared to other implementations not-based on the second eigenmode amplification. Additionally, a stiffer effective spring constant of the second eigenmode provides more stability of the oscillation during the surface scan.

Figure 4, shows schematics of the HD-KFM setup with an electric diagram of HD-KFM where the first flexural mode is excited mechanically and the second flexural mode is excited electrically. The topographic feedback operates on the mechanical amplitude (lock-in 1) and the electrical feedback operates on the electric amplitude (lock-in 2).

Figure 4. Schematics of HD-KFM with two lock-in.

BIMODAL HD-KFM. HIGH SENSITIVITY AND ROBUSTNESS

Graphene has emerged as a nanomaterial for the future due to its bidimensional structure and electronic properties. However, synthesis and transfer processes are not easy to implement especially on large areas where a single layer of graphene may have impurities or flakes of multiple layers of graphene. Although Raman spectroscopy is most commonly used to characterize monolayers of graphene, HD-KFM measurements have also proved to be useful to distinguish between single monolayer and multiple layers of graphene. Figure 5 shows both topography (Fig. 5a) and HD-KFM surface potential images (Fig 5b) of a graphene on Si sample. An ANSCM-PT probe (AppNano, USA) with a spring constant of 3 nN/nm was used. The imaging parameters were A01 = 18.5 nm, ASP1 = 18 nm, A02 = 1.9 nm, ASP1 = 1.6 nm. In the topographic image, there is an aggregate of several flakes which form an almost continuous film. Some areas show higher height values due to multiple layer stacks or self bending of the layers (white stripes in the image).  Simultaneously, the surface potential image (Fig 5b.) provides information of the layers with higher contrast.  Silicon oxides areas are depicted as red areas (with the colored-scale chosen). The graphene monolayers correspond to the blue areas, double layers to green areas and triple layers of graphene appear as orange areas. In addition, it can also be seen that some single monolayer areas in the bottom corner of the structure have slightly  higher potential values (notice the dark blue color) which can also indicate some surface charging of the SiO2 substrate below the graphene monolayer. Another channel typically used in resonant mode is the phase signal, which is related to energy dissipation mechanisms between tip and sample. Image 5c shows the corresponding phase signal channel obtained simultaneously to the surface potential, where no contrast among different layers is shown, only between substrate and graphene aggregate.

Figure 5. HD-KFM on graphene sample. Surface Potential measured with HD-KFM  (5b) allows to identify different layers of graphene much easier than topgraphy (5a and 5c).