Tomographic imaging using conductive atomic force microscopy

Sub-surface imaging properties of nanostructured materials have attracted significant research attention. These properties have been extensively explored using different modes of scanning probe microscopy, such as scanning thermal noise and scanning dielectric microscopy. These techniques heavily rely on the differences in the nanostructure material properties to achieve their functions. Consequently, there have been significant efforts devoted to developing tomography techniques using different approaches like scanning probe tomography. Unfortunately, most reported tomography techniques are generally destructive for the sample. Therefore, developing a non-destructive probe microscopy characterization technique for evaluating tomographic information is highly desirable.

Conductive atomic force microscopy (CAFM) is a type of AFM mode that is widely used in studying and characterizing defects in dielectric thin films as well as the electronic properties of nanocomposite films. Unlike other methods, CAFM can capture the current signals and surface topography simultaneously. This is achieved by scanning the sample surface in contact mode while using a conductive AFM probe to apply a voltage bias. Likewise, oxide probing via CAFM is generally based on tunneling electrical current through the sample. It is worth noting that CAFM is not surface sensitive and can effectively probe nanostructures embedded in an insulating material matrix.

“We viewed Co nanoparticles embedded inside the HfO₂ matrix of the granular films as essentially defects. So, we naturally thought it would be possible to image the granular film using CAFM. To our surprise, there were not prior work demonstrating imaging of granular films or nanostructures using CAFM in the scientific literature, and we became excited to try it for ourselves!”

With the advancement in nanotechnology, imaging and characterization of granular films have become extremely important. While different techniques, such as transmission electron microscopy (TEM), have been used to image these films, they fail to resolve nanocomposites with similar atomic masses. This has necessitated the development of effective alternative techniques for resolving the microstructure of granular films with poor contrast in TEM and other techniques. Motivated by the recent significant development in tomography imaging, Mr. Alexander Kang-Jun Toh and Professor Vivian Ng from the National University of Singapore developed an effective and reliable technique for studying granular films consisting of metallic grains embedded in an insulating matrix. Their work is currently published in the journal, Materials Characterization.

Briefly, the authors employed CAFM to image permalloy nanostructures embedded in Al₂O₃. Two permalloy nanostructure layers were fabricated using a combination of electron beam evaporation and nanosphere lithography. Bias-dependent CAFM imaging was modeled and investigated. The presented technique was extended to Co-HfO₂ granular films having randomly dispersed nanoparticles in the oxide matrix. The experimental findings were explained in detail using a phenomenological model based on direct tunneling, thermionic emission and Fowler-Nordheim tunneling.

“We were glad to find out that we indeed imaging different depths by changing the probing voltage of the CAFM, however, we were rather surprised to see that the bottom layer of the control sample was imaged first at a lower voltage. The understanding all came together with the model developed.”

The researchers demonstrated the bias dependence conduction of the top and bottom layers of the permalloy nanostructures. The isolated triangles and larger connected structures were formed by ideal nanosphere mask packing and defect patterns in the mask, respectively. The reconstruction of three-dimensional images of Co-HfO₂ granular film from CAFM images was illustrated. This further enabled non-destructive grain size analysis to obtain grain size information, which also agreed with the experimental data. Current-voltage modeling based on transmission coefficients obtained from thermionic emissions and tunneling models agreed well with the experimental data, suggesting the effectiveness and applicability of the CAFM imaging technique.

In summary, Toh and Ng successfully reported CAFM imaging of Co-HfO₂ granular films and stacked Py nanostructures as a function of applied bias and varying volume. The capability of CAFM on well-defined and patterned permalloy nanostructures was demonstrated. This non-destructive technique is relatively simple and cost-effective, making it a reliable alternative for obtaining grain size information from films and overcoming the issues and limitations associated with similar contrast often experienced in TEM techniques. In a statement to Advances in Engineering, Professor Vivian Ng explained that the new presented technique is a promising candidate for effective, reliable and precise tomographic imaging of nanostructures and devices.