In 1967 Müller and Panitz [Müller1968] combined the field ion microscope with a time of flight mass spectrometer. A small hole in the centre of the detector allowed atoms to enter the spectrometer. So a chemical analysis of selected areas of the specimen became possible. To achieve sufficient spatial resolution, early instruments had a very small aperture. Therefore, the total number of chemically analyzed atoms was only 0.1% of all evaporated atoms. Desirable would be a 100% identification rate of all incoming atoms, without losing the structural information.
Cerezo et al. presented a first prototype of a three dimensional atom probe (3DAP) in 1988 [Cerezo1988]. The replacement of the conventional detector system, consisting of a Micro-Channel-Plate (MCP) and phosphorous screen, by a position sensitive detector, made of a MCP and a wedge and strip anode, was a big step. Position data were collected from now on by a computer system.
Although the detector system was improved, the 3DAP is still based on the simple design of a FIM, but with unique differences. The focus in an atom probe measurement is of course on the specimen. Therefore, no image gas is used and only the field evaporated specimen atoms are analyzed.
The needle shaped specimen is still mounted in front of the detector system inside a vacuum chamber (Figure 2‑6). But additionally, an extraction electrode is added in front of the specimen tip. Small voltage pulses can be applied to the electrode to trigger the evaporation process. This is important for a correct time of flight measurement. In conventional instruments, a macroscopic electrode is used (2 - 4 mm in diameter), but in recent years, instruments with smaller extraction electrode, a so called local electrode, have been built [Kelly2004, Schlesiger2010]. The smaller extraction electrode leads to an increase in field strength of about a factor two and allows to operate with lower voltages or to measure a larger volume with the same voltage used in conventional systems. The use of laser pulses instead of high voltage pulses also allows to measure non conductive materials and helped to increase mass resolution. The development of time to digital converter (TDC) with higher time resolution (50 ps bin size) resulted in a reduction of flight length from 50 cm to around 15 cm, increasing the aperture of modern instrument and thus the measurable volume. From the gathered data the eroded specimen volume can be reconstructed with atomic resolution and chemical information.
The chemical nature of the evaporated atom is determined by simple flight time measurements. A short high voltage or laser pulse triggers the moment of evaporation and delivers a start point for the time of flight measurement. A high precision clock is started and stopped, when the detector registers an impact of a single atom. From the collected data the mass of the atom can be determined.
To calculate the mass from the time of flight it is assumed that the ion is instantaneous accelerated. This is a valid assumption, as the ion gains 90% of its kinetic energy within the first μm. But the total flight length is typically in the range of tenth of centimeters.
The unique feature of atom probe tomography is the reconstruction of the specimen volume from the obtained dataset, with atomic resolution and chemical information. This becomes possible, because for each evaporated atom the time of flight, the impact position on the detector and the respective voltages are stored. From this dataset, a three dimensional model has to be reconstructed. Different approaches are available for reconstruction [Blavette1995]. A reconstruction algorithm based on a point projection model is commonly used. The tip itself is modeled hemispherical. This assumption is not always valid and is object of current research on improved reconstruction algorithms. The x and y position within the sample are reconstructed from the impact positions xD and yD on the detector (Figure 2‑7), while the depth scale is calculated by the sequence of evaporated atoms. A constant volume is assigned to each individual atom species.
In the end a reconstructed specimen volume is obtained, as shown schematically in Figure 2-20 a). The chemical information is typically coded by color. From this volume, the determination of concentration profiles, interfaces and precipitates becomes possible. Besides one-dimensional composition profiles, which allow to determine a composition change across an interface, iso-concentration surfaces are a crucial tool to visualize and analyze structures within the three dimensional volume.
To obtain a iso-concentration surface, the volume is sampled by a small cube (Figure 2-10 b). The parameters of choice are the lateral dimensions of the cube and the overlap between adjoining cube positions (moving average).
From the atoms within a cube at a given position, a concentration value is calculated. Leading to a three dimensional concentration grid, where the overlap of the cubes defines the step size. The choice of the sampling size is crucial with regard to the error of of individual concentration values. Smaller cubes increase the error and therefore lower the reliability of the concentrations values. In contrast, cubes with larger dimensions, lead to an artificial broadening of the concentration profile.
By connecting concentration points of the same value, an iso-concentration profile can be obtained.