Types of Atom Probe

In 1967, Muller, Panitz and McLane developed the atom probe in order to identify the atoms that were imaged in the field ion microscope [5]. The original variant of the atom probe used a small hole in the center of the microchannel plate and phosphor screen assembly which served as the entrance aperture to a time-of-flight mass spectrometer, as shown in Fig. 2. The atoms were removed from the specimen at a well-defined time by the application of a short duration positive voltage pulse to the specimen. For many years, mechanical mercury-wetted reed relay based systems were used to produce these high voltage pulses at pulse repetition rates of between 50 and 200 Hz [1]. However, modern instruments now use solid-state devices that are capable of significantly higher repetition rates. The positive polarity high voltage pulse can be capacitively

Figure 1. Field ion micrographs of a) iridium, b) a Pd4oNi4oP2o bulk metallic glass, c) a grain boundary in NiAl, d) a dislocation in B2-ordered NiAl, e) a boron-decorated grain boundary in the heat affected zone of a molybdenum weldment and f) brightly-imaging yn and y' precipitates in Alloy 718. The dark circular region in the center of some of these images is dead region surrounding the entrance aperture of the mass spectrometer in the classical atom probe.

Figure 1. Field ion micrographs of a) iridium, b) a Pd4oNi4oP2o bulk metallic glass, c) a grain boundary in NiAl, d) a dislocation in B2-ordered NiAl, e) a boron-decorated grain boundary in the heat affected zone of a molybdenum weldment and f) brightly-imaging yn and y' precipitates in Alloy 718. The dark circular region in the center of some of these images is dead region surrounding the entrance aperture of the mass spectrometer in the classical atom probe.

Figure 2. Schematic diagrams of the classical atom probe, the three-dimensional atom probe and the local electrode atom probe. Some of the different types of detectors used on the three-dimensional atom probe are illustrated.

coupled to the standing voltage on the specimen. Alternatively and equivalently, a negative polarity pulse can be applied to a circular counter electrode a short distance in front of the specimen. The magnitude of the pulse voltage and the specimen temperature are selected to ensure that all the atoms present in the specimen are equally likely to field evaporate on the application of the high voltage pulse. Typical conditions are a specimen temperature 50—60 K for a wide variety of materials and ~20 K for aluminum alloys and a pulse voltage of 15 to 20% of the standing voltage. The correct parameters ensure that the more strongly bound elements are not retained on the surface and the less strongly bound elements are not preferentially removed by the standing voltage between the high voltage pulses thereby influencing the accuracy of the compositional determination. Calibration experiments are used to establish the correct conditions on new materials. A pulsed laser has also been used to momentarily heat the apex of the specimen to induce field evaporation on the standing voltage [6]. This pulsed laser method has been primarily used for the analysis of semiconducting specimens where the high voltage pulse would be attenuated before reaching the apex region of the specimen.

The field evaporated ions are detected on a single atom sensitive detector at the end of the mass spectrometer, as shown in Fig. 2. In the field evaporation process, the potential energy of the atom on the surface, neE, just prior to field evaporation is converted into kinetic energy, 1/2mv2 when the atom leaves the surface. Therefore, the mass-to-charge ratio, m/n, of each atom removed can be determined from mc n = dl (^c + Vpulse)t2, (1)

where d and t are the flight distance and flight time from the specimen to the detector, Vdc and Fpulse are the standing and pulse voltages on the specimen, and c is a constant used to convert the mass into atomic mass units. Although contrary to Muller's original concept of the operation of the atom probe, this process is generally conducted in the absence of the image gas to reduce the background noise. In the original atom probe, the flight time was manually measured from an oscilloscope trace. This method has been superseded with computer controlled high-speed digital timing systems. A typical sample in this type of atom probe is cylindrical volume with a diameter of up to ~2 nm that is defined by the entrance aperture and that contains approximately 10,000 to 50,000 atoms. Nowadays, this type of atom probe is referred to as a classical, conventional or one-dimensional atom probe.

The classical atom probe is very inefficient in the number of atoms collected from the specimen due to the small effective size of the aperture. Therefore, several variants have been developed to improve the collection efficiency. The 10 cm atom probe or imaging atom probe (IAP), designed by Panitz, had a significantly larger field of view [7,8]. This variant used a pair of spherically curved channel plates and phosphor screen so that the flight distance from the specimen to all points on the detector were identical. The detector could be momentarily energized to record only the ions of a given mass-to-charge ratio. Therefore, a two-dimensional map of the solute distribution could be obtained. However, these data were difficult to quantify and alternative approaches were developed.

These newer variants all feature an analysis chamber housing the field ion microscope and mass spectrometer, a specimen storage or preparation chamber and a fast entry air lock to load interchangeable specimens. The base pressure in the analysis chamber is achieved with either a turbomolecular, diffusion or ion pump supported with a titanium sublimation pump (TSP) or non-evaporable getter (NEG) pump and is typically less than 1 x 10-10 mbar. These designs differ primarily in the type ofposition-sensitive single atom detector that is used, as shown in Fig. 2. All the detectors are based on a stack of two (in a chevron configuration) or three (in a Z configuration) microchannel plates providing a gain of more than 106 times. However, different types of position-sensing elements are used. In these instruments, the active area of the detector together with the flight distance defines the area of analysis on the specimen's surface. Typical flight distances are smaller than the 1 to 2.5 m used in the one-dimensional atom probe and are typically between 0.45 and 0.62 m. These variants include the optical atom probe (OAP) that uses a phosphor screen and a CCD camera [9—11], the position-sensitive atom probe (PoSAP) that uses a three anode wedge-and-strip detector [12, 13], the tomographic atom probe (TAP) that uses a 10 x 10 square array of anodes [14], the optical position-sensitive atom probe that features a primary detector with a phosphor screen and a pair of secondary image-intensified detectors consisting of an 8 x 10 array of anodes and a phosphor screen and a CCD camera [15], and the optical tomo-graphic atom probe (OTAP) that uses a combination of a linearly segmented anode and a phosphor screen and a CCD camera [16]. These types of instruments are collectively referred to as three-dimensional atom probes (3DAP). A typical sampled volume in a three-dimensional atom probe has an area of ~10 x ~10 nm to ~20 x 20 nm and contains approximately 105 to 106 atoms. The typical time required to collect this number of atoms is of the order of 20 h. In both the classical atom probe and the three-dimensional atom probe, field ion microscopy is almost always performed at the start of the experiment. A region of the specimen is then selected for analysis by rotating the specimen about its apex until the field ion image of the selected region covers either the probe aperture or the single atom detector. The image gas is then removed from the system and the specimen is field evaporated until a sufficient number of atoms are collected or the specimen fails.

A new variant of three-dimensional atom probe known as a scanning atom probe (SAP) or a local electrode atom probe (LEAP®) has been introduced recently [17—20]. This instrument uses a small aperture or local electrode positioned extremely close to the apex of the specimen, as shown in Fig. 2c. A photograph of the local electrode atom probe is shown in Fig. 3. Nishikawa's original concept of the scanning atom probe was that the local electrode could be scanned across the surface of a rough specimen and a natural protrusion could be selected for analysis thereby extending the type of specimen that could be studied. However, the analysis of natural protrusions is limited to simple compositional analysis since the non-uniform end forms of these types of protrusions do not permit reliable reconstructions of the positions of the atoms. In addition, surface contamination and surface diffusion can have a serious influence on the reliability of the analysis. It is also possible to analyze traditional needle-shaped field ion specimens with this variant. In addition, in-situ analyses may be performed on many individual microtips in two-dimensional arrays of microtips that are fabricated from the surface of flat specimens by focused ion beam based techniques. An integral part of this instrument is a nano-positioning piezoelectric stage to accurately align

Figure 3. Photograph of the local electrode atom probe. Courtesy Imago Scientific Corporation.

the aperture in the local electrode with either the apex of a traditional needle-shaped specimen or a protrusion. The alignment of the specimen with the local electrode is performed while viewing the process with a pair of orthogonal high resolution video cameras attached to long working distance microscopes.

The primary advantage of this local electrode configuration is that the voltage that has to be applied to the specimen is significantly reduced from traditional atom probe configurations. For example, only ~50% of the voltage is required for a 30 |lm aperture positioned ~10 |m in front of the apex of the specimen compared to the 5—10 mm diameter apertures placed 4—10 mm in front of the specimen in the conventional three-dimensional atom probe. This lower voltage not only extends the range of the instrument to blunter specimens but, more importantly, allows the instrument to be operated at significantly faster repetition rates that are possible due to the correspondingly lower pulse voltage. The repetition rate currently available commercially for generating these high voltage pulses is 200 kHz and the limit is likely to be in excess of 1 MHz. In order to take advantage of these high repetition rates, a different type of position-sensitive detector is used. Therefore, the local electrode atom probe uses a chevron microchannel plate with a crossed delay line coupled to an ultrafast digital timing system. Unlike other types of atom probe, the close proximity of the local electrode with the specimen dictates that the design uses a vibration isolated cryo-generator to minimize any change in the field due to small changes in the alignment of specimen to the local electrode. This local electrode configuration also enables the single atom detector to be positioned significantly closer to the specimen without the normally associated degradation in mass resolution. As this configuration has a wide field of view, the design and operation of the local electrode atom probe can also be simplified by eliminating the goniometer that is normally used to rotate the specimen to select the region of analysis. A typical sample volume in the local electrode atom probe contains 106 to 3 x 107 atoms and datasets containing over 108 atoms have been collected. Due to the faster high voltage pulser, detector and electronics, the analysis time has been reduced by several orders of magnitude (~300X) compared to the traditional three-dimensional atom probe. Typical collection rates in the local electrode atom probe are 5 to 10 x 106 ions per hour. This high collection rate enables many specimens to be characterized in a day.

Since a time-of-flight mass spectrometer is used in all of these instruments, the atom probe is able to detect and has equal sensitivity for all elements. Unlike some other techniques, no pre-selection of elemental species is required to perform an analysis. The mass resolution of the atom probe is limited by small deficits in the energy of the ions. These energy deficits arise due to the full magnitude of the pulse not being transferred to the ion as it leaves the specimen. The initial method to improve the mass resolution in the one-dimensional atom probe was the addition of a Poschenreider lens to the mass spectrometer [2, 21]. This lens is a section of a toroid with matched length field free regions. Ions with different energies take slightly different length paths in the lens and are isochronally focused on the single atom detector. One additional advantage of the Poschenreider lens is that it can filter out any residual image gas and other impurity ions that are field evaporated on the standing voltage before they strike the detector thereby improving the background noise level in the analyses. Unfortunately, the Poschenreider lens configuration cannot be used in the three-dimensional atom probes due to their large acceptance angles. Therefore, reflectron lenses or energy mirrors have been used in three-dimensional atom probes [2, 22]. In this type of lens, the ions enter a region of increasing field and are reflected back to the detector. Ions with lower energy do not travel as far into the lens as ions with the full energy and are isochronally focused on the detector. However, the reflectron also reduces the detection efficiency by ~10% due to a field defining mesh at the entrance/exit of the lens. These energy-compensation devices improve the mass resolution so that the individual isotopes of the elements can be resolved. Due to the close proximity of the counter electrode, energy compensation is not required in the local electrode atom probe to resolve the individual isotopes. However, further improvement in the mass resolution can be achieved with a method based on post acceleration [23]. This post acceleration method can also increase the field of view.

2.3. Specimen Preparation

As with many techniques, the quality of the specimen is one of the most important parameters in the success of an experiment. The aim of specimen preparation is to produce a needle-shaped specimen with an end radius that is less than approximately 50 nm. In addition, the taper angle ofthe shank ofthe needle should be no larger than 10° as this parameter defines the maximum number of atoms that may be collected from the specimen. Most atom probe specimens are produced by either electropolishing or ion milling techniques depending on the material, original shape and the type of specimen [1, 2].

In most metallic specimens, the bulk sample is reduced into square or cylindrical shaped bars that are ~0.25 mm across by ~10 mm long usually with the use of a diamond wire saw or diamond impregnated cutting wheel. Other starting forms of material such as wires or whiskers are also suitable. The bar is generally crimped into an annealed copper tube to facilitate handling in subsequent stages. Electropolishing is generally performed in two stages. In the first stage, the specimen is suspended in a 5 to 6-mm-thick layer of electrolyte floated on top of a dense inert liquid such as a polyfluorinated polyether. The electropolishing cell is configured with the specimen as the anode and a wire or circular cathode made of gold or platinum. Since only the middle portion of the specimen is in the electrolyte, a necked region is formed in the central region of the bar. For some materials, the electropolishing process is continued until the necked region is too narrow to support the weight of the bottom half of the bar and two atom probe specimens are made. However for many materials, the process is normally terminated before this point is reached so that a second stage with more controlled electropolishing conditions may be used. Since the necked region has been formed in the first stage, the specimen may be transferred to a simpler cell without the inert liquid and electropolishing continued until separation occurs. A list of suitable electrolytes for a wide variety of materials may be found elsewhere [2]. The resulting needles are then carefully cleaned in a suitable solvent to remove any traces of the electrolyte. Specimens are generally examined in an optical microscope at magnifications of 100 to 200X to check the quality of the surface finish, taper angle and the end radius before insertion into the atom probe.

Specimens that are either too blunt for further analysis or had fractured during analysis may be resharpened. This micropolishing method may also be used on freshly made or corroded specimens to remove any damaged regions or surface films. In this method, the specimen (anode) is carefully positioned in a drop ofelectrolyte suspended in a platinum loop (cathode). The electrolyte is held in place in the ~3-mm-diameter loop by surface tension. A voltage is then applied to the cell to electropolish the region of the specimen in the electrolyte. The position of the specimen in the electrolyte may be adjusted with the use of a micromanipulator and is generally monitored under a low power (~30X) microscope. A desired taper angle may be sculpted by adjusting position and the time spent polishing in the electrolyte.

Thin film specimens such as surface, bi- or multi-layer films are generally fabricated into the needle-shaped specimens with the use of a focused ion beam (FIB) miller.

The aim of the fabrication process is to position the area of interest in the apex of the needle. One common method is to use a silicon substrate that has been Bosch etched into an array of ~5 x ~5 |lm square or rod posts [1, 24]. The resistivity of the silicon should be less than 0.05 Q-cm to ensure that the pulse voltage is not attenuated in the specimen and thereby rendered ineffective. Alternatively, posts may be cut with the use of a dicing saw [25]. The multilayer film is then deposited on the surfaces of these posts. A 2- to 3-|m-thick platinum cap is deposited on top of the region of interest, i.e. the center of the post, to protect the underlying region from gallium implantation during milling. In specimens without this cap, concentrations of up to 30% gallium has been measured. The post is then ion milled with an annular 30 keV gallium ion beam where the outer diameter is slightly larger than the extent of the post and the inner diameter is ~2 |m. The inner diameter is progressively reduced to ~0.06 |m during milling to produce the desired taper angle and end radius. Milling is terminated when the end radius of the specimen is less than ~50 nm and the last remnant of the platinum cap is removed. Examination or monitoring of the specimen should be performed with the electron beam rather than the gallium ion beam to minimize gallium implantation, and therefore instruments featuring ion and electron columns within a single system are preferred.

In the case of the local electrode atom probe, this milling method may also be used to fabricate specimens from a specific location on the surface of a specimen by placing a platinum cap or marker at the desired location and then milling a wide moat around that cap before preparing the microtip as described above. This milling method may also be used to fabricate specimens of fine powders. For this starting form of material, an individual powder particle is attached to the end of a relatively blunt needle with either a conducting epoxy or a platinum bridge deposited in the FIB.

Was this article helpful?

0 0

Post a comment