Uncertainties in focused ion beam characterisation

Focused ion beam (FIB) and dual-beam microscopy is a common tool used for materials characterisation, including 3D cross-sectional imaging and micro-mechanical analysis. These techniques have enabled new and revolutionary insight into the behaviour of materials at the microscale, however relating these properties to the bulk material to produce useful quantitative data is more challenging. With a wide range of instruments and setups available, as well as user variability, it is very difficult to control consistency and quality of FIB acquired data. This project examines some of the artefacts produced in materials by the act of characterisation with focused ion beam, causing uncertainties and errors in the acquired data and subsequent measurements on the dataset. Fundamental experiments on silicon-based samples are done initially because it is a well-characterised and simple material, which behaves predictably with FIB milling. Tungsten carbide cobalt (WC/Co) hardmetals are then studied because they are sensitive to FIB milling, undergoing microstructural changes and having a very slow milling rate, making them a good study material. The artefacts investigated are: the variation in slice thickness during serial sectioning for 3D reconstruction, phase transformation in the sample material, and non-perfect geometries and ion-induced damage in micro-mechanical test specimens, specifically micropillars. The experiments carried out demonstrate the severity of each artefact and develop possible prevention methods and improvements to the methods. Slice thickness variation in serial sectioning cannot be directly prevented as it is inherent to the instrument. However, by including a measurement device with known geometry in the acquisition of the data, the slices can be measured during data processing and the values appropriately adjusted in volume reconstructions. The development of the invention and analysis processes of such a device is described. In an attempt to standardise slice thickness measurement, a possible solution for mass production of the devices is also discussed. Phase transformation during milling causes uncertainty in microstructural characterisation of a susceptible material, due to the induced phase disguising the existing phase composition of the material. FIB milling does not affect all materials in this way, but one of the most severe occurrences is in cobalt and the cobalt binder phase in WC/Co hardmetals. The underlying mechanism of the transformation is explored and the preparation conditions that avoid it are identified. Micro-mechanical testing is a newer implemented capability of FIB microscopy, and one of the most commonly tested geometries is micro-pillars for compression testing. Due to the shape and damaging nature of the beam, uncertainties associated with reproducibility of micro-pillar specimens and induced changes in the material can affect the mechanical properties of the material being tested. This has significant ramifications on the reliability and validity of quantitative measurements made on micro-pillars, especially when comparing with bulk material properties. A few experiments are described that investigate different aspects of micro-pillar experiments, including a focus on metrology (using silicon), feasibility when applied to a complex composite material (WC/Co), and practical and interesting aspects when applied to a single crystal material (WC)