Overview #
Cross-sections allow one to look inside the bulk sample and to reveal sample features which are hidden below the sample surface. FIB-SEMs have the unique capability to section and analyze samples at precisely selected points which cannot be achieved by another technique. Figure 1 displays a cross section of an array of nanopores. The cross-section allows visualization the structure of the pores below the sample surface.
The Ga and plasma FIB-SEMs are conventionally used for cross-sectioning. The area of interest for the cross-section is identified using the electron beam. A protective layer can deposited to protect the surface using a gas injection system (GIS) and either electron- or ion-beam induced deposition. Common gases used for protective layers are: tungsten, platinum, or carbon. It is important to choose a gas that matches the hardness of the sample. The gas deposits a smooth surface to reduce curtaining and protect the sample surface from the ion beam during the cross-sectioning process. A cross-section is then cut with the ion beam to expose the sub surface feature. The cross-section is finished off by polishing the cross-section surface with a smaller ion beam current.
Detailed Cross-Section Process #
Choosing the area of interest #
In this first step, the area of interest is identified using the SEM (if available). Features or interest can be identified using the secondary electron (SE), backscattered electron (BSE), energy dispersive (EDS), or backscattered electron diffraction (EBSD) detectors. Features of interest can be anything: cracks homogeneous materials; interfaces between different materials in geological samples; the areas with specific elements, phases or crystallographic orientation/structure; multilayered stack of lithographically and epitaxially defined layers in semiconductor devices (CMOS); etc. One may have or may not have knowledge or information about the expected features to be exposed by cross-sectioning. The area of interested may be close to the surface or deeply below the surface.
The area of interest is typically tilted so that the surface of the sample is perpendicular to the ion beam (Figure 2-1). Smaller tilts can also be performed to expose tilted features, or to elongate the appearance of interfaces between layers.
Adding a protection layer #
To prevent the misinterpretations of the milling results and to protect the sample from the erosion while performing the most aggressive ion milling, a protective layer is deposited on top of the feature of interest which will add thickness to the sample. The material(s) for the deposition can be chosen based on the configuration of gas injection system (GIS) installed on your FIB. A simple recommendation is to use a material for beam induced deposition (BID) which will give you the best contrast against the surface for your analysis.
Typically, an electron-BID (EBID) protecting layer is deposited in a rectangular pattern to cover the area of interest with some margins (depending on the sample). This step is used to protect the very top surface of the sample from ion-beam damage. If the area of interest in the cross-section is deeper within the structure, this step can be omitted.
Patterning of a thicker patch of material then is done in a smaller size to protect just the area above planned cross-section, allowing several iterations of polishing and considering the error of exact localization of the site of interest (Figure 2-2). The thicker section is typically deposited using ion-beam induced deposition (IBID) because of the larger deposition rates with the ion beam relative to the electron beam.
Cross-Section Milling Considerations #
Milling Pattern #
The material to be removed to expose the cross-section face is milled in a triangular trench to minimize the amount of material that is removed while exposing the entire section face (Figure 2-1 and 2-3). The triangle is created by increasing the amount of ion exposure closer to the cross-section face, increasing the amount of material that is sputtered. In some FIB-SEM systems, this type of pattern is called a rough cross-section. For bulk removal, a small gap is typically left between the milling pattern and the area of interest which is removed in the polishing step.
Cross-section Size #
The geometry of the cut is set by the size of the feature that will be exposed. Typically, the depth of the trench is roughly twice as deep as the size of the feature of interest, while the length of the trench is determined by the angle between the electron and ion beams (θ). The minimum length of trench needed to have unobscured imaging of the cross-section is the cross-section depth × tan(θ), which is about 1.3X the depth for systems with 52° between electron and ion beams. It is typically suggested to increase the length of the pattern beyond the minimum length to ensure exposure of the feature of interest.
Scan Direction #
The scan direction is important to avoid re-deposition on the cross-section area. The scan direction is set to mill towards the deposition layer resulting in re-deposition that will occur away from the cross-section and not cause problems.
Milling Current #
Choosing the appropriate ion beam current will enable fast cross-sectioning, which should take ~20min or less. Of course, the current chosen is sample dependent because not all materials can handle high beam currents and therefore require longer cross-sectioning times. The rough suggestions in Table 1 intended as a starting guide. It is important to adjust the beam current appropriately to the sample and can mean doing several test runs to check for artifacts and optimize for milling time. The available beam currents may also differ on different FIB systems and it is recommended to check the user manual or with the vendor about appropriate guide lines for each system.
| Cross-Section Depth | Suggested Beam Current | Notes |
|---|---|---|
| < 5 μm | 1 nA | |
| <15 μm | 15 nA | |
| < 50 μm | > 30 nA | |
| < 100 μm | > 100 nA | Typically on available with PFIBs |
| > 100 μm | > 2 μA | Typically on available with PFIBs |
EDS and EBSD Analysis #
Specific geometries for EDS and EBSD cross-sectional analysis are required to be cut and are instrument specific and depend on where the detectors are located within the FIB-SEM system. Check with the instrument vendor about the specific requirements for the system in operation.
Polishing #
Polish the cross-section surface using a cleaning cross-section pattern/polishing pattern. This is done with a reduced beam current (1/2 to 1/5 of the beam current that was used for cross-sectioning). The smaller final probe diameter means that a smoother cut can be made and the cross-section surface will be polished, induced artefacts such as curtaining which increase for increased beam currents will be reduced.
Note: if curtaining is a problem (which can be more pronounced for PFIBs), a rocking stage can be used in the polishing process.
Analysis #
After polishing is complete, analysis of the section can be performed by a variety of methods: imaging, spectroscopy, etc. Some things to keep in mind when performing analysis:
- The sample will likely be tilted so images will be foreshortened and need to corrected for the tilt
- Areas at the sides of the cross-section may be shadowed from detectors and may produce less detectable signal relative to areas at the center of the cross-section
- There will be some contamination of the cross-section surface with the milling element, which can cause other issues (see milling artifacts).
